* [PATCH 0/6] ARM: oprofile: use perf-events framework as backend [v2] @ 2010-03-10 10:41 Will Deacon 2010-03-10 10:41 ` [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c Will Deacon 0 siblings, 1 reply; 19+ messages in thread From: Will Deacon @ 2010-03-10 10:41 UTC (permalink / raw) To: linux-arm-kernel This is version 2 of the patchset originally posted here: http://lists.infradead.org/pipermail/linux-arm-kernel/2010-February/010476.html The only changes from the original code concern the PMU ID scheme and name resolution. This is now handled using an enum of IDs rather than a list of #defines. Patches taken against 2.6.33. Cc: Russell King - ARM Linux <linux@arm.linux.org.uk> Cc: Ingo Molnar <mingo@elte.hu> Cc: Jamie Iles <jamie.iles@picochip.com> Cc: Jean Pihet <jpihet@mvista.com> Will Deacon (6): ARM: perf-events: add Realview PMU IRQs to pmu.c ARM: perf-events: use numeric ID to identify PMU ARM: perf-events: add support for xscale PMUs perf-events: export enable/disable event symbols to kernel modules ARM: oprofile: use perf-events framework as backend ARM: oprofile: remove old files and update KConfig arch/arm/Kconfig | 26 +- arch/arm/include/asm/perf_event.h | 14 + arch/arm/kernel/perf_event.c | 862 ++++++++++++++++++++++++++++++- arch/arm/kernel/pmu.c | 14 + arch/arm/oprofile/Makefile | 7 +- arch/arm/oprofile/backtrace.c | 83 --- arch/arm/oprofile/common.c | 356 +++++++++++-- arch/arm/oprofile/op_arm_model.h | 35 -- arch/arm/oprofile/op_counter.h | 27 - arch/arm/oprofile/op_model_arm11_core.c | 162 ------ arch/arm/oprofile/op_model_arm11_core.h | 45 -- arch/arm/oprofile/op_model_mpcore.c | 306 ----------- arch/arm/oprofile/op_model_mpcore.h | 61 --- arch/arm/oprofile/op_model_v6.c | 78 --- arch/arm/oprofile/op_model_v7.c | 415 --------------- arch/arm/oprofile/op_model_v7.h | 103 ---- arch/arm/oprofile/op_model_xscale.c | 444 ---------------- include/linux/perf_event.h | 1 + kernel/perf_event.c | 8 + 19 files changed, 1194 insertions(+), 1853 deletions(-) delete mode 100644 arch/arm/oprofile/backtrace.c delete mode 100644 arch/arm/oprofile/op_arm_model.h delete mode 100644 arch/arm/oprofile/op_counter.h delete mode 100644 arch/arm/oprofile/op_model_arm11_core.c delete mode 100644 arch/arm/oprofile/op_model_arm11_core.h delete mode 100644 arch/arm/oprofile/op_model_mpcore.c delete mode 100644 arch/arm/oprofile/op_model_mpcore.h delete mode 100644 arch/arm/oprofile/op_model_v6.c delete mode 100644 arch/arm/oprofile/op_model_v7.c delete mode 100644 arch/arm/oprofile/op_model_v7.h delete mode 100644 arch/arm/oprofile/op_model_xscale.c ^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c 2010-03-10 10:41 [PATCH 0/6] ARM: oprofile: use perf-events framework as backend [v2] Will Deacon @ 2010-03-10 10:41 ` Will Deacon 2010-03-10 10:41 ` [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU Will Deacon 2010-03-10 21:59 ` [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c Tony Lindgren 0 siblings, 2 replies; 19+ messages in thread From: Will Deacon @ 2010-03-10 10:41 UTC (permalink / raw) To: linux-arm-kernel Add the PMU IRQs for the Realview boards to kernel/pmu.c so that they can be used by the perf-events framework. Signed-off-by: Will Deacon <will.deacon@arm.com> --- arch/arm/kernel/pmu.c | 14 ++++++++++++++ 1 files changed, 14 insertions(+), 0 deletions(-) diff --git a/arch/arm/kernel/pmu.c b/arch/arm/kernel/pmu.c index a124312..11a321e 100644 --- a/arch/arm/kernel/pmu.c +++ b/arch/arm/kernel/pmu.c @@ -36,6 +36,20 @@ static const int irqs[] = { IRQ_EB11MP_PMU_CPU1, IRQ_EB11MP_PMU_CPU2, IRQ_EB11MP_PMU_CPU3, +#elif defined(CONFIG_MACH_REALVIEW_PB1176) + IRQ_DC1176_CORE_PMU, +#elif defined(CONFIG_MACH_REALVIEW_PB11MP) + IRQ_TC11MP_PMU_CPU0, + IRQ_TC11MP_PMU_CPU1, + IRQ_TC11MP_PMU_CPU2, + IRQ_TC11MP_PMU_CPU3, +#elif defined(CONFIG_MACH_REALVIEW_PBA8) + IRQ_PBA8_PMU, +#elif defined(CONFIG_MACH_REALVIEW_PBX) + IRQ_PBX_PMU_CPU0, + IRQ_PBX_PMU_CPU1, + IRQ_PBX_PMU_CPU2, + IRQ_PBX_PMU_CPU3, #elif defined(CONFIG_ARCH_OMAP3) INT_34XX_BENCH_MPU_EMUL, #elif defined(CONFIG_ARCH_IOP32X) -- 1.6.3.3 ^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU 2010-03-10 10:41 ` [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c Will Deacon @ 2010-03-10 10:41 ` Will Deacon 2010-03-10 10:41 ` [PATCH 3/6] ARM: perf-events: add support for xscale PMUs Will Deacon 2010-03-10 11:03 ` [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU Jamie Iles 2010-03-10 21:59 ` [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c Tony Lindgren 1 sibling, 2 replies; 19+ messages in thread From: Will Deacon @ 2010-03-10 10:41 UTC (permalink / raw) To: linux-arm-kernel The ARM perf-events framework provides support for a number of different PMUs using struct arm_pmu. The char *name field of this struct can be used to identify the PMU, but this is cumbersome if used outside of perf. This patch replaces the name string for a PMU with an enum, which holds a unique ID for the PMU being represented. This ID can be used to index an array of names within perf, so no functionality is lost. The presence of the ID field, allows other kernel subsystems [currently oprofile] to use their own mappings for the PMU name. Cc: Jamie Iles <jamie.iles@picochip.com> Cc: Jean Pihet <jpihet@mvista.com> Signed-off-by: Will Deacon <will.deacon@arm.com> --- arch/arm/include/asm/perf_event.h | 14 +++++++++++++ arch/arm/kernel/perf_event.c | 39 +++++++++++++++++++++++++++--------- 2 files changed, 43 insertions(+), 10 deletions(-) diff --git a/arch/arm/include/asm/perf_event.h b/arch/arm/include/asm/perf_event.h index 49e3049..fa4b326 100644 --- a/arch/arm/include/asm/perf_event.h +++ b/arch/arm/include/asm/perf_event.h @@ -28,4 +28,18 @@ set_perf_event_pending(void) * same indexes here for consistency. */ #define PERF_EVENT_INDEX_OFFSET 1 +/* ARM perf PMU IDs for use by internal perf clients. */ +enum arm_perf_pmu_ids { + ARM_PERF_PMU_ID_XSCALE1 = 0, + ARM_PERF_PMU_ID_XSCALE2, + ARM_PERF_PMU_ID_V6, + ARM_PERF_PMU_ID_V6MP, + ARM_PERF_PMU_ID_CA8, + ARM_PERF_PMU_ID_CA9, + ARM_NUM_PMU_IDS, +}; + +extern enum arm_perf_pmu_ids +armpmu_get_pmu_id(void); + #endif /* __ARM_PERF_EVENT_H__ */ diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c index c54ceb3..e1a5f5f 100644 --- a/arch/arm/kernel/perf_event.c +++ b/arch/arm/kernel/perf_event.c @@ -16,6 +16,7 @@ #include <linux/interrupt.h> #include <linux/kernel.h> +#include <linux/module.h> #include <linux/perf_event.h> #include <linux/spinlock.h> #include <linux/uaccess.h> @@ -67,8 +68,18 @@ struct cpu_hw_events { }; DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events); +/* PMU names. */ +static const char *arm_pmu_names[] = { + [ARM_PERF_PMU_ID_XSCALE1] = "xscale1", + [ARM_PERF_PMU_ID_XSCALE2] = "xscale2", + [ARM_PERF_PMU_ID_V6] = "v6", + [ARM_PERF_PMU_ID_V6MP] = "v6mpcore", + [ARM_PERF_PMU_ID_CA8] = "ARMv7 Cortex-A8", + [ARM_PERF_PMU_ID_CA9] = "ARMv7 Cortex-A9", +}; + struct arm_pmu { - char *name; + enum arm_perf_pmu_ids id; irqreturn_t (*handle_irq)(int irq_num, void *dev); void (*enable)(struct hw_perf_event *evt, int idx); void (*disable)(struct hw_perf_event *evt, int idx); @@ -87,6 +98,18 @@ struct arm_pmu { /* Set at runtime when we know what CPU type we are. */ static const struct arm_pmu *armpmu; +enum arm_perf_pmu_ids +armpmu_get_pmu_id(void) +{ + int id = -ENODEV; + + if (armpmu != NULL) + id = armpmu->id; + + return id; +} +EXPORT_SYMBOL_GPL(armpmu_get_pmu_id); + #define HW_OP_UNSUPPORTED 0xFFFF #define C(_x) \ @@ -1143,7 +1166,7 @@ armv6mpcore_pmu_disable_event(struct hw_perf_event *hwc, } static const struct arm_pmu armv6pmu = { - .name = "v6", + .id = ARM_PERF_PMU_ID_V6, .handle_irq = armv6pmu_handle_irq, .enable = armv6pmu_enable_event, .disable = armv6pmu_disable_event, @@ -1166,7 +1189,7 @@ static const struct arm_pmu armv6pmu = { * reset the period and enable the interrupt reporting. */ static const struct arm_pmu armv6mpcore_pmu = { - .name = "v6mpcore", + .id = ARM_PERF_PMU_ID_V6MP, .handle_irq = armv6pmu_handle_irq, .enable = armv6pmu_enable_event, .disable = armv6mpcore_pmu_disable_event, @@ -1196,10 +1219,6 @@ static const struct arm_pmu armv6mpcore_pmu = { * counter and all 4 performance counters together can be reset separately. */ -#define ARMV7_PMU_CORTEX_A8_NAME "ARMv7 Cortex-A8" - -#define ARMV7_PMU_CORTEX_A9_NAME "ARMv7 Cortex-A9" - /* Common ARMv7 event types */ enum armv7_perf_types { ARMV7_PERFCTR_PMNC_SW_INCR = 0x00, @@ -2104,7 +2123,7 @@ init_hw_perf_events(void) perf_max_events = armv6mpcore_pmu.num_events; break; case 0xC080: /* Cortex-A8 */ - armv7pmu.name = ARMV7_PMU_CORTEX_A8_NAME; + armv7pmu.id = ARM_PERF_PMU_ID_CA8; memcpy(armpmu_perf_cache_map, armv7_a8_perf_cache_map, sizeof(armv7_a8_perf_cache_map)); armv7pmu.event_map = armv7_a8_pmu_event_map; @@ -2116,7 +2135,7 @@ init_hw_perf_events(void) perf_max_events = armv7pmu.num_events; break; case 0xC090: /* Cortex-A9 */ - armv7pmu.name = ARMV7_PMU_CORTEX_A9_NAME; + armv7pmu.id = ARM_PERF_PMU_ID_CA9; memcpy(armpmu_perf_cache_map, armv7_a9_perf_cache_map, sizeof(armv7_a9_perf_cache_map)); armv7pmu.event_map = armv7_a9_pmu_event_map; @@ -2135,7 +2154,7 @@ init_hw_perf_events(void) if (armpmu) pr_info("enabled with %s PMU driver, %d counters available\n", - armpmu->name, armpmu->num_events); + arm_pmu_names[armpmu->id], armpmu->num_events); return 0; } -- 1.6.3.3 ^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 3/6] ARM: perf-events: add support for xscale PMUs 2010-03-10 10:41 ` [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU Will Deacon @ 2010-03-10 10:41 ` Will Deacon 2010-03-10 10:41 ` [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules Will Deacon 2010-03-10 11:03 ` [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU Jamie Iles 1 sibling, 1 reply; 19+ messages in thread From: Will Deacon @ 2010-03-10 10:41 UTC (permalink / raw) To: linux-arm-kernel The perf-events framework for ARM only supports v6 and v7 cores. This patch adds support for xscale v1 and v2 PMUs to perf, based on the OProfile drivers in arch/arm/oprofile/op_model_xscale.c Signed-off-by: Will Deacon <will.deacon@arm.com> --- arch/arm/kernel/perf_event.c | 825 +++++++++++++++++++++++++++++++++++++++++- 1 files changed, 819 insertions(+), 6 deletions(-) diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c index e1a5f5f..cf014dc 100644 --- a/arch/arm/kernel/perf_event.c +++ b/arch/arm/kernel/perf_event.c @@ -2097,6 +2097,801 @@ static u32 __init armv7_reset_read_pmnc(void) return nb_cnt + 1; } +/* + * ARMv5 [xscale] Performance counter handling code. + * + * Based on xscale OProfile code. + * + * There are two variants of the xscale PMU that we support: + * - xscale1pmu: 2 event counters and a cycle counter + * - xscale2pmu: 4 event counters and a cycle counter + * The two variants share event definitions, but have different + * PMU structures. + */ + +enum xscale_perf_types { + XSCALE_PERFCTR_ICACHE_MISS = 0x00, + XSCALE_PERFCTR_ICACHE_NO_DELIVER = 0x01, + XSCALE_PERFCTR_DATA_STALL = 0x02, + XSCALE_PERFCTR_ITLB_MISS = 0x03, + XSCALE_PERFCTR_DTLB_MISS = 0x04, + XSCALE_PERFCTR_BRANCH = 0x05, + XSCALE_PERFCTR_BRANCH_MISS = 0x06, + XSCALE_PERFCTR_INSTRUCTION = 0x07, + XSCALE_PERFCTR_DCACHE_FULL_STALL = 0x08, + XSCALE_PERFCTR_DCACHE_FULL_STALL_CONTIG = 0x09, + XSCALE_PERFCTR_DCACHE_ACCESS = 0x0A, + XSCALE_PERFCTR_DCACHE_MISS = 0x0B, + XSCALE_PERFCTR_DCACE_WRITE_BACK = 0x0C, + XSCALE_PERFCTR_PC_CHANGED = 0x0D, + XSCALE_PERFCTR_BCU_REQUEST = 0x10, + XSCALE_PERFCTR_BCU_FULL = 0x11, + XSCALE_PERFCTR_BCU_DRAIN = 0x12, + XSCALE_PERFCTR_BCU_ECC_NO_ELOG = 0x14, + XSCALE_PERFCTR_BCU_1_BIT_ERR = 0x15, + XSCALE_PERFCTR_RMW = 0x16, + /* XSCALE_PERFCTR_CCNT is not hardware defined */ + XSCALE_PERFCTR_CCNT = 0xFE, + XSCALE_PERFCTR_UNUSED = 0xFF, +}; + +enum xscale_counters { + XSCALE_CYCLE_COUNTER = 1, + XSCALE_COUNTER0, + XSCALE_COUNTER1, + XSCALE_COUNTER2, + XSCALE_COUNTER3, +}; + +static const unsigned xscale_perf_map[PERF_COUNT_HW_MAX] = { + [PERF_COUNT_HW_CPU_CYCLES] = XSCALE_PERFCTR_CCNT, + [PERF_COUNT_HW_INSTRUCTIONS] = XSCALE_PERFCTR_INSTRUCTION, + [PERF_COUNT_HW_CACHE_REFERENCES] = HW_OP_UNSUPPORTED, + [PERF_COUNT_HW_CACHE_MISSES] = HW_OP_UNSUPPORTED, + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = XSCALE_PERFCTR_BRANCH, + [PERF_COUNT_HW_BRANCH_MISSES] = XSCALE_PERFCTR_BRANCH_MISS, + [PERF_COUNT_HW_BUS_CYCLES] = HW_OP_UNSUPPORTED, +}; + +static const unsigned xscale_perf_cache_map[PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX] = { + [C(L1D)] = { + [C(OP_READ)] = { + [C(RESULT_ACCESS)] = XSCALE_PERFCTR_DCACHE_ACCESS, + [C(RESULT_MISS)] = XSCALE_PERFCTR_DCACHE_MISS, + }, + [C(OP_WRITE)] = { + [C(RESULT_ACCESS)] = XSCALE_PERFCTR_DCACHE_ACCESS, + [C(RESULT_MISS)] = XSCALE_PERFCTR_DCACHE_MISS, + }, + [C(OP_PREFETCH)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + }, + [C(L1I)] = { + [C(OP_READ)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = XSCALE_PERFCTR_ICACHE_MISS, + }, + [C(OP_WRITE)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = XSCALE_PERFCTR_ICACHE_MISS, + }, + [C(OP_PREFETCH)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + }, + [C(LL)] = { + [C(OP_READ)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + [C(OP_WRITE)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + [C(OP_PREFETCH)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + }, + [C(DTLB)] = { + [C(OP_READ)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = XSCALE_PERFCTR_DTLB_MISS, + }, + [C(OP_WRITE)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = XSCALE_PERFCTR_DTLB_MISS, + }, + [C(OP_PREFETCH)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + }, + [C(ITLB)] = { + [C(OP_READ)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = XSCALE_PERFCTR_ITLB_MISS, + }, + [C(OP_WRITE)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = XSCALE_PERFCTR_ITLB_MISS, + }, + [C(OP_PREFETCH)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + }, + [C(BPU)] = { + [C(OP_READ)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + [C(OP_WRITE)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + [C(OP_PREFETCH)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + }, +}; + +#define XSCALE_PMU_ENABLE 0x001 +#define XSCALE_PMN_RESET 0x002 +#define XSCALE_CCNT_RESET 0x004 +#define XSCALE_PMU_RESET (CCNT_RESET | PMN_RESET) +#define XSCALE_PMU_CNT64 0x008 + +static inline int +xscalepmu_event_map(int config) +{ + int mapping = xscale_perf_map[config]; + if (HW_OP_UNSUPPORTED == mapping) + mapping = -EOPNOTSUPP; + return mapping; +} + +static u64 +xscalepmu_raw_event(u64 config) +{ + return config & 0xff; +} + +#define XSCALE1_OVERFLOWED_MASK 0x700 +#define XSCALE1_CCOUNT_OVERFLOW 0x400 +#define XSCALE1_COUNT0_OVERFLOW 0x100 +#define XSCALE1_COUNT1_OVERFLOW 0x200 +#define XSCALE1_CCOUNT_INT_EN 0x040 +#define XSCALE1_COUNT0_INT_EN 0x010 +#define XSCALE1_COUNT1_INT_EN 0x020 +#define XSCALE1_COUNT0_EVT_SHFT 12 +#define XSCALE1_COUNT0_EVT_MASK (0xff << XSCALE1_COUNT0_EVT_SHFT) +#define XSCALE1_COUNT1_EVT_SHFT 20 +#define XSCALE1_COUNT1_EVT_MASK (0xff << XSCALE1_COUNT1_EVT_SHFT) + +static inline u32 +xscale1pmu_read_pmnc(void) +{ + u32 val; + asm volatile("mrc p14, 0, %0, c0, c0, 0" : "=r" (val)); + return val; +} + +static inline void +xscale1pmu_write_pmnc(u32 val) +{ + /* upper 4bits and 7, 11 are write-as-0 */ + val &= 0xffff77f; + asm volatile("mcr p14, 0, %0, c0, c0, 0" : : "r" (val)); +} + +static inline int +xscale1_pmnc_counter_has_overflowed(unsigned long pmnc, + enum xscale_counters counter) +{ + int ret = 0; + + switch (counter) { + case XSCALE_CYCLE_COUNTER: + ret = pmnc & XSCALE1_CCOUNT_OVERFLOW; + break; + case XSCALE_COUNTER0: + ret = pmnc & XSCALE1_COUNT0_OVERFLOW; + break; + case XSCALE_COUNTER1: + ret = pmnc & XSCALE1_COUNT1_OVERFLOW; + break; + default: + WARN_ONCE(1, "invalid counter number (%d)\n", counter); + } + + return ret; +} + +static irqreturn_t +xscale1pmu_handle_irq(int irq_num, void *dev) +{ + unsigned long pmnc; + struct perf_sample_data data; + struct cpu_hw_events *cpuc; + struct pt_regs *regs; + int idx; + + /* + * NOTE: there's an A stepping erratum that states if an overflow + * bit already exists and another occurs, the previous + * Overflow bit gets cleared. There's no workaround. + * Fixed in B stepping or later. + */ + pmnc = xscale1pmu_read_pmnc(); + + /* + * Write the value back to clear the overflow flags. Overflow + * flags remain in pmnc for use below. We also disable the PMU + * while we process the interrupt. + */ + xscale1pmu_write_pmnc(pmnc & ~XSCALE_PMU_ENABLE); + + if (!(pmnc & XSCALE1_OVERFLOWED_MASK)) + return IRQ_NONE; + + regs = get_irq_regs(); + data.addr = 0; + + cpuc = &__get_cpu_var(cpu_hw_events); + for (idx = 0; idx <= armpmu->num_events; ++idx) { + struct perf_event *event = cpuc->events[idx]; + struct hw_perf_event *hwc; + + if (!test_bit(idx, cpuc->active_mask)) + continue; + + if (!xscale1_pmnc_counter_has_overflowed(pmnc, idx)) + continue; + + hwc = &event->hw; + armpmu_event_update(event, hwc, idx); + data.period = event->hw.last_period; + if (!armpmu_event_set_period(event, hwc, idx)) + continue; + + if (perf_event_overflow(event, 0, &data, regs)) + armpmu->disable(hwc, idx); + } + + perf_event_do_pending(); + + /* + * Re-enable the PMU. + */ + pmnc = xscale1pmu_read_pmnc() | XSCALE_PMU_ENABLE; + xscale1pmu_write_pmnc(pmnc); + + return IRQ_HANDLED; +} + +static void +xscale1pmu_enable_event(struct hw_perf_event *hwc, int idx) +{ + unsigned long val, mask, evt, flags; + + switch (idx) { + case XSCALE_CYCLE_COUNTER: + mask = 0; + evt = XSCALE1_CCOUNT_INT_EN; + break; + case XSCALE_COUNTER0: + mask = XSCALE1_COUNT0_EVT_MASK; + evt = (hwc->config_base << XSCALE1_COUNT0_EVT_SHFT) | + XSCALE1_COUNT0_INT_EN; + break; + case XSCALE_COUNTER1: + mask = XSCALE1_COUNT1_EVT_MASK; + evt = (hwc->config_base << XSCALE1_COUNT1_EVT_SHFT) | + XSCALE1_COUNT1_INT_EN; + break; + default: + WARN_ONCE(1, "invalid counter number (%d)\n", idx); + return; + } + + spin_lock_irqsave(&pmu_lock, flags); + val = xscale1pmu_read_pmnc(); + val &= ~mask; + val |= evt; + xscale1pmu_write_pmnc(val); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static void +xscale1pmu_disable_event(struct hw_perf_event *hwc, int idx) +{ + unsigned long val, mask, evt, flags; + + switch (idx) { + case XSCALE_CYCLE_COUNTER: + mask = XSCALE1_CCOUNT_INT_EN; + evt = 0; + break; + case XSCALE_COUNTER0: + mask = XSCALE1_COUNT0_INT_EN | XSCALE1_COUNT0_EVT_MASK; + evt = XSCALE_PERFCTR_UNUSED << XSCALE1_COUNT0_EVT_SHFT; + break; + case XSCALE_COUNTER1: + mask = XSCALE1_COUNT1_INT_EN | XSCALE1_COUNT1_EVT_MASK; + evt = XSCALE_PERFCTR_UNUSED << XSCALE1_COUNT1_EVT_SHFT; + break; + default: + WARN_ONCE(1, "invalid counter number (%d)\n", idx); + return; + } + + spin_lock_irqsave(&pmu_lock, flags); + val = xscale1pmu_read_pmnc(); + val &= ~mask; + val |= evt; + xscale1pmu_write_pmnc(val); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static int +xscale1pmu_get_event_idx(struct cpu_hw_events *cpuc, + struct hw_perf_event *event) +{ + if (XSCALE_PERFCTR_CCNT == event->config_base) { + if (test_and_set_bit(XSCALE_CYCLE_COUNTER, cpuc->used_mask)) + return -EAGAIN; + + return XSCALE_CYCLE_COUNTER; + } else { + if (!test_and_set_bit(XSCALE_COUNTER1, cpuc->used_mask)) { + return XSCALE_COUNTER1; + } + + if (!test_and_set_bit(XSCALE_COUNTER0, cpuc->used_mask)) { + return XSCALE_COUNTER0; + } + + return -EAGAIN; + } +} + +static void +xscale1pmu_start(void) +{ + unsigned long flags, val; + + spin_lock_irqsave(&pmu_lock, flags); + val = xscale1pmu_read_pmnc(); + val |= XSCALE_PMU_ENABLE; + xscale1pmu_write_pmnc(val); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static void +xscale1pmu_stop(void) +{ + unsigned long flags, val; + + spin_lock_irqsave(&pmu_lock, flags); + val = xscale1pmu_read_pmnc(); + val &= ~XSCALE_PMU_ENABLE; + xscale1pmu_write_pmnc(val); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static inline u32 +xscale1pmu_read_counter(int counter) +{ + u32 val = 0; + + switch (counter) { + case XSCALE_CYCLE_COUNTER: + asm volatile("mrc p14, 0, %0, c1, c0, 0" : "=r" (val)); + break; + case XSCALE_COUNTER0: + asm volatile("mrc p14, 0, %0, c2, c0, 0" : "=r" (val)); + break; + case XSCALE_COUNTER1: + asm volatile("mrc p14, 0, %0, c3, c0, 0" : "=r" (val)); + break; + } + + return val; +} + +static inline void +xscale1pmu_write_counter(int counter, u32 val) +{ + switch (counter) { + case XSCALE_CYCLE_COUNTER: + asm volatile("mcr p14, 0, %0, c1, c0, 0" : : "r" (val)); + break; + case XSCALE_COUNTER0: + asm volatile("mcr p14, 0, %0, c2, c0, 0" : : "r" (val)); + break; + case XSCALE_COUNTER1: + asm volatile("mcr p14, 0, %0, c3, c0, 0" : : "r" (val)); + break; + } +} + +static const struct arm_pmu xscale1pmu = { + .id = ARM_PERF_PMU_ID_XSCALE1, + .handle_irq = xscale1pmu_handle_irq, + .enable = xscale1pmu_enable_event, + .disable = xscale1pmu_disable_event, + .event_map = xscalepmu_event_map, + .raw_event = xscalepmu_raw_event, + .read_counter = xscale1pmu_read_counter, + .write_counter = xscale1pmu_write_counter, + .get_event_idx = xscale1pmu_get_event_idx, + .start = xscale1pmu_start, + .stop = xscale1pmu_stop, + .num_events = 3, + .max_period = (1LLU << 32) - 1, +}; + +#define XSCALE2_OVERFLOWED_MASK 0x01f +#define XSCALE2_CCOUNT_OVERFLOW 0x001 +#define XSCALE2_COUNT0_OVERFLOW 0x002 +#define XSCALE2_COUNT1_OVERFLOW 0x004 +#define XSCALE2_COUNT2_OVERFLOW 0x008 +#define XSCALE2_COUNT3_OVERFLOW 0x010 +#define XSCALE2_CCOUNT_INT_EN 0x001 +#define XSCALE2_COUNT0_INT_EN 0x002 +#define XSCALE2_COUNT1_INT_EN 0x004 +#define XSCALE2_COUNT2_INT_EN 0x008 +#define XSCALE2_COUNT3_INT_EN 0x010 +#define XSCALE2_COUNT0_EVT_SHFT 0 +#define XSCALE2_COUNT0_EVT_MASK (0xff << XSCALE2_COUNT0_EVT_SHFT) +#define XSCALE2_COUNT1_EVT_SHFT 8 +#define XSCALE2_COUNT1_EVT_MASK (0xff << XSCALE2_COUNT1_EVT_SHFT) +#define XSCALE2_COUNT2_EVT_SHFT 16 +#define XSCALE2_COUNT2_EVT_MASK (0xff << XSCALE2_COUNT2_EVT_SHFT) +#define XSCALE2_COUNT3_EVT_SHFT 24 +#define XSCALE2_COUNT3_EVT_MASK (0xff << XSCALE2_COUNT3_EVT_SHFT) + +static inline u32 +xscale2pmu_read_pmnc(void) +{ + u32 val; + asm volatile("mrc p14, 0, %0, c0, c1, 0" : "=r" (val)); + /* bits 1-2 and 4-23 are read-unpredictable */ + return val & 0xff000009; +} + +static inline void +xscale2pmu_write_pmnc(u32 val) +{ + /* bits 4-23 are write-as-0, 24-31 are write ignored */ + val &= 0xf; + asm volatile("mcr p14, 0, %0, c0, c1, 0" : : "r" (val)); +} + +static inline u32 +xscale2pmu_read_overflow_flags(void) +{ + u32 val; + asm volatile("mrc p14, 0, %0, c5, c1, 0" : "=r" (val)); + return val; +} + +static inline void +xscale2pmu_write_overflow_flags(u32 val) +{ + asm volatile("mcr p14, 0, %0, c5, c1, 0" : : "r" (val)); +} + +static inline u32 +xscale2pmu_read_event_select(void) +{ + u32 val; + asm volatile("mrc p14, 0, %0, c8, c1, 0" : "=r" (val)); + return val; +} + +static inline void +xscale2pmu_write_event_select(u32 val) +{ + asm volatile("mcr p14, 0, %0, c8, c1, 0" : : "r"(val)); +} + +static inline u32 +xscale2pmu_read_int_enable(void) +{ + u32 val; + asm volatile("mrc p14, 0, %0, c4, c1, 0" : "=r" (val)); + return val; +} + +static void +xscale2pmu_write_int_enable(u32 val) +{ + asm volatile("mcr p14, 0, %0, c4, c1, 0" : : "r" (val)); +} + +static inline int +xscale2_pmnc_counter_has_overflowed(unsigned long of_flags, + enum xscale_counters counter) +{ + int ret = 0; + + switch (counter) { + case XSCALE_CYCLE_COUNTER: + ret = of_flags & XSCALE2_CCOUNT_OVERFLOW; + break; + case XSCALE_COUNTER0: + ret = of_flags & XSCALE2_COUNT0_OVERFLOW; + break; + case XSCALE_COUNTER1: + ret = of_flags & XSCALE2_COUNT1_OVERFLOW; + break; + case XSCALE_COUNTER2: + ret = of_flags & XSCALE2_COUNT2_OVERFLOW; + break; + case XSCALE_COUNTER3: + ret = of_flags & XSCALE2_COUNT3_OVERFLOW; + break; + default: + WARN_ONCE(1, "invalid counter number (%d)\n", counter); + } + + return ret; +} + +static irqreturn_t +xscale2pmu_handle_irq(int irq_num, void *dev) +{ + unsigned long pmnc, of_flags; + struct perf_sample_data data; + struct cpu_hw_events *cpuc; + struct pt_regs *regs; + int idx; + + /* Disable the PMU. */ + pmnc = xscale2pmu_read_pmnc(); + xscale2pmu_write_pmnc(pmnc & ~XSCALE_PMU_ENABLE); + + /* Check the overflow flag register. */ + of_flags = xscale2pmu_read_overflow_flags(); + if (!(of_flags & XSCALE2_OVERFLOWED_MASK)) + return IRQ_NONE; + + /* Clear the overflow bits. */ + xscale2pmu_write_overflow_flags(of_flags); + + regs = get_irq_regs(); + data.addr = 0; + + cpuc = &__get_cpu_var(cpu_hw_events); + for (idx = 0; idx <= armpmu->num_events; ++idx) { + struct perf_event *event = cpuc->events[idx]; + struct hw_perf_event *hwc; + + if (!test_bit(idx, cpuc->active_mask)) + continue; + + if (!xscale2_pmnc_counter_has_overflowed(pmnc, idx)) + continue; + + hwc = &event->hw; + armpmu_event_update(event, hwc, idx); + data.period = event->hw.last_period; + if (!armpmu_event_set_period(event, hwc, idx)) + continue; + + if (perf_event_overflow(event, 0, &data, regs)) + armpmu->disable(hwc, idx); + } + + perf_event_do_pending(); + + /* + * Re-enable the PMU. + */ + pmnc = xscale2pmu_read_pmnc() | XSCALE_PMU_ENABLE; + xscale2pmu_write_pmnc(pmnc); + + return IRQ_HANDLED; +} + +static void +xscale2pmu_enable_event(struct hw_perf_event *hwc, int idx) +{ + unsigned long flags, ien, evtsel; + + ien = xscale2pmu_read_int_enable(); + evtsel = xscale2pmu_read_event_select(); + + switch (idx) { + case XSCALE_CYCLE_COUNTER: + ien |= XSCALE2_CCOUNT_INT_EN; + break; + case XSCALE_COUNTER0: + ien |= XSCALE2_COUNT0_INT_EN; + evtsel &= ~XSCALE2_COUNT0_EVT_MASK; + evtsel |= hwc->config_base << XSCALE2_COUNT0_EVT_SHFT; + break; + case XSCALE_COUNTER1: + ien |= XSCALE2_COUNT1_INT_EN; + evtsel &= ~XSCALE2_COUNT1_EVT_MASK; + evtsel |= hwc->config_base << XSCALE2_COUNT1_EVT_SHFT; + break; + case XSCALE_COUNTER2: + ien |= XSCALE2_COUNT2_INT_EN; + evtsel &= ~XSCALE2_COUNT2_EVT_MASK; + evtsel |= hwc->config_base << XSCALE2_COUNT2_EVT_SHFT; + break; + case XSCALE_COUNTER3: + ien |= XSCALE2_COUNT3_INT_EN; + evtsel &= ~XSCALE2_COUNT3_EVT_MASK; + evtsel |= hwc->config_base << XSCALE2_COUNT3_EVT_SHFT; + break; + default: + WARN_ONCE(1, "invalid counter number (%d)\n", idx); + return; + } + + spin_lock_irqsave(&pmu_lock, flags); + xscale2pmu_write_event_select(evtsel); + xscale2pmu_write_int_enable(ien); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static void +xscale2pmu_disable_event(struct hw_perf_event *hwc, int idx) +{ + unsigned long flags, ien, evtsel; + + ien = xscale2pmu_read_int_enable(); + evtsel = xscale2pmu_read_event_select(); + + switch (idx) { + case XSCALE_CYCLE_COUNTER: + ien &= ~XSCALE2_CCOUNT_INT_EN; + break; + case XSCALE_COUNTER0: + ien &= ~XSCALE2_COUNT0_INT_EN; + evtsel &= ~XSCALE2_COUNT0_EVT_MASK; + evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT0_EVT_SHFT; + break; + case XSCALE_COUNTER1: + ien &= ~XSCALE2_COUNT1_INT_EN; + evtsel &= ~XSCALE2_COUNT1_EVT_MASK; + evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT1_EVT_SHFT; + break; + case XSCALE_COUNTER2: + ien &= ~XSCALE2_COUNT2_INT_EN; + evtsel &= ~XSCALE2_COUNT2_EVT_MASK; + evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT2_EVT_SHFT; + break; + case XSCALE_COUNTER3: + ien &= ~XSCALE2_COUNT3_INT_EN; + evtsel &= ~XSCALE2_COUNT3_EVT_MASK; + evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT3_EVT_SHFT; + break; + default: + WARN_ONCE(1, "invalid counter number (%d)\n", idx); + return; + } + + spin_lock_irqsave(&pmu_lock, flags); + xscale2pmu_write_event_select(evtsel); + xscale2pmu_write_int_enable(ien); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static int +xscale2pmu_get_event_idx(struct cpu_hw_events *cpuc, + struct hw_perf_event *event) +{ + int idx = xscale1pmu_get_event_idx(cpuc, event); + if (idx >= 0) + goto out; + + if (!test_and_set_bit(XSCALE_COUNTER3, cpuc->used_mask)) + idx = XSCALE_COUNTER3; + else if (!test_and_set_bit(XSCALE_COUNTER2, cpuc->used_mask)) + idx = XSCALE_COUNTER2; +out: + return idx; +} + +static void +xscale2pmu_start(void) +{ + unsigned long flags, val; + + spin_lock_irqsave(&pmu_lock, flags); + val = xscale2pmu_read_pmnc() & ~XSCALE_PMU_CNT64; + val |= XSCALE_PMU_ENABLE; + xscale2pmu_write_pmnc(val); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static void +xscale2pmu_stop(void) +{ + unsigned long flags, val; + + spin_lock_irqsave(&pmu_lock, flags); + val = xscale2pmu_read_pmnc(); + val &= ~XSCALE_PMU_ENABLE; + xscale2pmu_write_pmnc(val); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static inline u32 +xscale2pmu_read_counter(int counter) +{ + u32 val = 0; + + switch (counter) { + case XSCALE_CYCLE_COUNTER: + asm volatile("mrc p14, 0, %0, c1, c1, 0" : "=r" (val)); + break; + case XSCALE_COUNTER0: + asm volatile("mrc p14, 0, %0, c0, c2, 0" : "=r" (val)); + break; + case XSCALE_COUNTER1: + asm volatile("mrc p14, 0, %0, c1, c2, 0" : "=r" (val)); + break; + case XSCALE_COUNTER2: + asm volatile("mrc p14, 0, %0, c2, c2, 0" : "=r" (val)); + break; + case XSCALE_COUNTER3: + asm volatile("mrc p14, 0, %0, c3, c2, 0" : "=r" (val)); + break; + } + + return val; +} + +static inline void +xscale2pmu_write_counter(int counter, u32 val) +{ + switch (counter) { + case XSCALE_CYCLE_COUNTER: + asm volatile("mcr p14, 0, %0, c1, c1, 0" : : "r" (val)); + break; + case XSCALE_COUNTER0: + asm volatile("mcr p14, 0, %0, c0, c2, 0" : : "r" (val)); + break; + case XSCALE_COUNTER1: + asm volatile("mcr p14, 0, %0, c1, c2, 0" : : "r" (val)); + break; + case XSCALE_COUNTER2: + asm volatile("mcr p14, 0, %0, c2, c2, 0" : : "r" (val)); + break; + case XSCALE_COUNTER3: + asm volatile("mcr p14, 0, %0, c3, c2, 0" : : "r" (val)); + break; + } +} + +static const struct arm_pmu xscale2pmu = { + .id = ARM_PERF_PMU_ID_XSCALE2, + .handle_irq = xscale2pmu_handle_irq, + .enable = xscale2pmu_enable_event, + .disable = xscale2pmu_disable_event, + .event_map = xscalepmu_event_map, + .raw_event = xscalepmu_raw_event, + .read_counter = xscale2pmu_read_counter, + .write_counter = xscale2pmu_write_counter, + .get_event_idx = xscale2pmu_get_event_idx, + .start = xscale2pmu_start, + .stop = xscale2pmu_stop, + .num_events = 5, + .max_period = (1LLU << 32) - 1, +}; + static int __init init_hw_perf_events(void) { @@ -2104,7 +2899,7 @@ init_hw_perf_events(void) unsigned long implementor = (cpuid & 0xFF000000) >> 24; unsigned long part_number = (cpuid & 0xFFF0); - /* We only support ARM CPUs implemented by ARM at the moment. */ + /* ARM Ltd CPUs. */ if (0x41 == implementor) { switch (part_number) { case 0xB360: /* ARM1136 */ @@ -2146,15 +2941,33 @@ init_hw_perf_events(void) armv7pmu.num_events = armv7_reset_read_pmnc(); perf_max_events = armv7pmu.num_events; break; - default: - pr_info("no hardware support available\n"); - perf_max_events = -1; + } + /* Intel CPUs [xscale]. */ + } else if (0x69 == implementor) { + part_number = (cpuid >> 13) & 0x7; + switch (part_number) { + case 1: + armpmu = &xscale1pmu; + memcpy(armpmu_perf_cache_map, xscale_perf_cache_map, + sizeof(xscale_perf_cache_map)); + perf_max_events = xscale1pmu.num_events; + break; + case 2: + armpmu = &xscale2pmu; + memcpy(armpmu_perf_cache_map, xscale_perf_cache_map, + sizeof(xscale_perf_cache_map)); + perf_max_events = xscale2pmu.num_events; + break; } } - if (armpmu) + if (armpmu) { pr_info("enabled with %s PMU driver, %d counters available\n", - arm_pmu_names[armpmu->id], armpmu->num_events); + arm_pmu_names[armpmu->id], armpmu->num_events); + } else { + pr_info("no hardware support available\n"); + perf_max_events = -1; + } return 0; } -- 1.6.3.3 ^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules 2010-03-10 10:41 ` [PATCH 3/6] ARM: perf-events: add support for xscale PMUs Will Deacon @ 2010-03-10 10:41 ` Will Deacon 2010-03-10 10:41 ` [PATCH 5/6] ARM: oprofile: use perf-events framework as backend Will Deacon 2010-03-10 10:56 ` [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules Peter Zijlstra 0 siblings, 2 replies; 19+ messages in thread From: Will Deacon @ 2010-03-10 10:41 UTC (permalink / raw) To: linux-arm-kernel The perf_event_enable and perf_event_disable functions are used to control the activation of perf-events for a given context or CPU. This patch exports these symbols so that they can be used by kernel modules. Without these symbols, an event must be destroyed and recreated to disable or enable it respectively. The maximum number of perf-events is also made available to modules via the perf_get_max_events function. Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Frederic Weisbecker <fweisbec@gmail.com> Signed-off-by: Will Deacon <will.deacon@arm.com> --- include/linux/perf_event.h | 1 + kernel/perf_event.c | 8 ++++++++ 2 files changed, 9 insertions(+), 0 deletions(-) diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index a177698..c28a0ee 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -741,6 +741,7 @@ struct perf_output_handle { * Set by architecture code: */ extern int perf_max_events; +extern int perf_get_max_events(void); extern const struct pmu *hw_perf_event_init(struct perf_event *event); diff --git a/kernel/perf_event.c b/kernel/perf_event.c index 2ae7409..0157033 100644 --- a/kernel/perf_event.c +++ b/kernel/perf_event.c @@ -39,6 +39,12 @@ static DEFINE_PER_CPU(struct perf_cpu_context, perf_cpu_context); int perf_max_events __read_mostly = 1; +int perf_get_max_events(void) +{ + return perf_max_events; +} +EXPORT_SYMBOL_GPL(perf_get_max_events); + static int perf_reserved_percpu __read_mostly; static int perf_overcommit __read_mostly = 1; @@ -604,6 +610,7 @@ void perf_event_disable(struct perf_event *event) raw_spin_unlock_irq(&ctx->lock); } +EXPORT_SYMBOL_GPL(perf_event_disable); static int event_sched_in(struct perf_event *event, @@ -1028,6 +1035,7 @@ void perf_event_enable(struct perf_event *event) out: raw_spin_unlock_irq(&ctx->lock); } +EXPORT_SYMBOL_GPL(perf_event_enable); static int perf_event_refresh(struct perf_event *event, int refresh) { -- 1.6.3.3 ^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 5/6] ARM: oprofile: use perf-events framework as backend 2010-03-10 10:41 ` [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules Will Deacon @ 2010-03-10 10:41 ` Will Deacon 2010-03-10 10:41 ` [PATCH 6/6] ARM: oprofile: remove old files and update KConfig Will Deacon 2010-03-10 10:56 ` [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules Peter Zijlstra 1 sibling, 1 reply; 19+ messages in thread From: Will Deacon @ 2010-03-10 10:41 UTC (permalink / raw) To: linux-arm-kernel There are currently two hardware performance monitoring subsystems in the kernel for ARM: OProfile and perf-events. This creates the following problems: 1.) Duplicate PMU accessor code. Inevitable code drift may lead to bugs in one framework that are fixed in the other. 2.) Locking issues. OProfile doesn't reprogram hardware counters between profiling runs if the events to be monitored have not been changed. This means that other profiling frameworks cannot use the counters if OProfile is in use. 3.) Due to differences in the two frameworks, it may not be possible to compare the results obtained by OProfile with those obtained by perf. This patch removes the OProfile PMU driver code and replaces it with calls to perf, therefore solving the issues mentioned above. The only userspace-visible change is the lack of SCU counter support for 11MPCore. This is currently unsupported by OProfile userspace tools anyway and therefore shouldn't cause any problems. Signed-off-by: Will Deacon <will.deacon@arm.com> --- arch/arm/oprofile/common.c | 356 ++++++++++++++++++++++++++++++++++++++------ 1 files changed, 307 insertions(+), 49 deletions(-) diff --git a/arch/arm/oprofile/common.c b/arch/arm/oprofile/common.c index 3fcd752..c61d476 100644 --- a/arch/arm/oprofile/common.c +++ b/arch/arm/oprofile/common.c @@ -2,32 +2,210 @@ * @file common.c * * @remark Copyright 2004 Oprofile Authors + * @remark Copyright 2010 ARM Ltd. * @remark Read the file COPYING * * @author Zwane Mwaikambo + * @author Will Deacon [move to perf] */ +#include <linux/cpumask.h> +#include <linux/errno.h> #include <linux/init.h> +#include <linux/mutex.h> #include <linux/oprofile.h> -#include <linux/errno.h> +#include <linux/perf_event.h> #include <linux/slab.h> #include <linux/sysdev.h> -#include <linux/mutex.h> +#include <asm/stacktrace.h> +#include <linux/uaccess.h> -#include "op_counter.h" -#include "op_arm_model.h" +#include <asm/perf_event.h> +#include <asm/ptrace.h> + +#ifdef CONFIG_HW_PERF_EVENTS +/* + * Per performance monitor configuration as set via oprofilefs. + */ +struct op_counter_config { + unsigned long count; + unsigned long enabled; + unsigned long event; + unsigned long unit_mask; + unsigned long kernel; + unsigned long user; + struct perf_event_attr attr; +}; -static struct op_arm_model_spec *op_arm_model; static int op_arm_enabled; static DEFINE_MUTEX(op_arm_mutex); -struct op_counter_config *counter_config; +static struct op_counter_config *counter_config; +static struct perf_event **perf_events[nr_cpumask_bits]; +static int perf_num_counters; + +/* + * Create perf attributes to mirror the oprofile settings in counter_config. + * Attributes are created in the disabled state and are permanently scheduled + * on the PMU. + */ +static void op_setup_counter_attrs(void) +{ + int i; + u32 size = sizeof(struct perf_event_attr); + struct perf_event_attr *attr; + + for (i = 0; i < perf_num_counters; ++i) { + attr = &counter_config[i].attr; + memset(attr, 0, size); + attr->type = PERF_TYPE_RAW; + attr->size = size; + attr->config = counter_config[i].event; + attr->sample_period = counter_config[i].count; + attr->pinned = 1; + attr->disabled = 1; + } +} + +/* + * Overflow callback for oprofile. + */ +static void op_overflow_handler(struct perf_event *event, int unused, + struct perf_sample_data *data, struct pt_regs *regs) +{ + int id; + u32 cpu = smp_processor_id(); + + for (id = 0; id < perf_num_counters; ++id) + if (perf_events[cpu][id] == event) + break; + + if (id != perf_num_counters) + oprofile_add_sample(regs, id); + else + pr_warning("oprofile: ignoring spurious overflow " + "on cpu %u\n", cpu); +} + +/* + * Create a new perf event for event `id' on cpu `cpu' using the latest + * perf attribute for the corresponding counter. + */ +static int op_update_perf_event(int cpu, int id) +{ + int ret = 0; + struct perf_event *event = perf_events[cpu][id]; + struct perf_event_attr *attr = &counter_config[id].attr; + + if (event != NULL) + perf_event_release_kernel(event); + + event = perf_event_create_kernel_counter(attr, cpu, -1, + op_overflow_handler); + if (IS_ERR(event)) { + ret = PTR_ERR(event); + event = NULL; + } + + perf_events[cpu][id] = event; + return ret; +} + +/* + * Called by op_arm_setup to program the PMU with the parameters held in + * counter_config. + */ +static int op_perf_setup(void) +{ + int cpu, event, ret = 0; + + op_setup_counter_attrs(); + + for_each_online_cpu(cpu) { + for (event = 0; event < perf_num_counters; ++event) { + ret = op_update_perf_event(cpu, event); + if (ret) + break; + } + } + + return ret; +} + +/* + * Enable the perf events whose counter_configs are also enabled. + * We check that the event becomes active because we specify that + * the attribute is pinned. + */ +static int op_enable_counter(int cpu, int event) +{ + int ret = 0; + + if (counter_config[event].enabled) { + perf_event_enable(perf_events[cpu][event]); + if (perf_events[cpu][event]->state != PERF_EVENT_STATE_ACTIVE) { + pr_warning("oprofile: failed to enable event %d " + "on CPU %d\n", event, cpu); + ret = -EBUSY; + } + } + + return ret; +} + +static void op_perf_stop(void) +{ + int cpu, event; + + for_each_online_cpu(cpu) { + for (event = 0; event < perf_num_counters; ++event) + perf_event_disable(perf_events[cpu][event]); + } +} + +static int op_perf_start(void) +{ + int cpu, event, ret = 0; + + for_each_online_cpu(cpu) { + for (event = 0; event < perf_num_counters; ++event) { + ret = op_enable_counter(cpu, event); + if (ret) { + /* Disabling a disabled event is idempotent. */ + op_perf_stop(); + break; + } + } + } + + return ret; +} + +static char *op_name_from_perf_id(enum arm_perf_pmu_ids id) +{ + switch (id) { + case ARM_PERF_PMU_ID_XSCALE1: + return "arm/xscale1"; + case ARM_PERF_PMU_ID_XSCALE2: + return "arm/xscale2"; + case ARM_PERF_PMU_ID_V6: + return "arm/armv6"; + case ARM_PERF_PMU_ID_V6MP: + return "arm/mpcore"; + case ARM_PERF_PMU_ID_CA8: + return "arm/armv7"; + case ARM_PERF_PMU_ID_CA9: + return "arm/armv7-ca9"; + default: + return NULL; + } +} static int op_arm_create_files(struct super_block *sb, struct dentry *root) { unsigned int i; - for (i = 0; i < op_arm_model->num_counters; i++) { + for (i = 0; i < perf_num_counters; i++) { struct dentry *dir; char buf[4]; @@ -49,7 +227,7 @@ static int op_arm_setup(void) int ret; spin_lock(&oprofilefs_lock); - ret = op_arm_model->setup_ctrs(); + ret = op_perf_setup(); spin_unlock(&oprofilefs_lock); return ret; } @@ -60,8 +238,9 @@ static int op_arm_start(void) mutex_lock(&op_arm_mutex); if (!op_arm_enabled) { - ret = op_arm_model->start(); - op_arm_enabled = !ret; + ret = 0; + op_perf_start(); + op_arm_enabled = 1; } mutex_unlock(&op_arm_mutex); return ret; @@ -71,7 +250,7 @@ static void op_arm_stop(void) { mutex_lock(&op_arm_mutex); if (op_arm_enabled) - op_arm_model->stop(); + op_perf_stop(); op_arm_enabled = 0; mutex_unlock(&op_arm_mutex); } @@ -81,7 +260,7 @@ static int op_arm_suspend(struct sys_device *dev, pm_message_t state) { mutex_lock(&op_arm_mutex); if (op_arm_enabled) - op_arm_model->stop(); + op_perf_stop(); mutex_unlock(&op_arm_mutex); return 0; } @@ -89,7 +268,7 @@ static int op_arm_suspend(struct sys_device *dev, pm_message_t state) static int op_arm_resume(struct sys_device *dev) { mutex_lock(&op_arm_mutex); - if (op_arm_enabled && op_arm_model->start()) + if (op_arm_enabled && op_perf_start()) op_arm_enabled = 0; mutex_unlock(&op_arm_mutex); return 0; @@ -126,58 +305,137 @@ static void exit_driverfs(void) #define exit_driverfs() do { } while (0) #endif /* CONFIG_PM */ -int __init oprofile_arch_init(struct oprofile_operations *ops) +static int report_trace(struct stackframe *frame, void *d) { - struct op_arm_model_spec *spec = NULL; - int ret = -ENODEV; + unsigned int *depth = d; - ops->backtrace = arm_backtrace; + if (*depth) { + oprofile_add_trace(frame->pc); + (*depth)--; + } + + return *depth == 0; +} -#ifdef CONFIG_CPU_XSCALE - spec = &op_xscale_spec; -#endif +/* + * The registers we're interested in are at the end of the variable + * length saved register structure. The fp points at the end of this + * structure so the address of this struct is: + * (struct frame_tail *)(xxx->fp)-1 + */ +struct frame_tail { + struct frame_tail *fp; + unsigned long sp; + unsigned long lr; +} __attribute__((packed)); -#ifdef CONFIG_OPROFILE_ARMV6 - spec = &op_armv6_spec; -#endif +static struct frame_tail* user_backtrace(struct frame_tail *tail) +{ + struct frame_tail buftail[2]; -#ifdef CONFIG_OPROFILE_MPCORE - spec = &op_mpcore_spec; -#endif + /* Also check accessibility of one struct frame_tail beyond */ + if (!access_ok(VERIFY_READ, tail, sizeof(buftail))) + return NULL; + if (__copy_from_user_inatomic(buftail, tail, sizeof(buftail))) + return NULL; -#ifdef CONFIG_OPROFILE_ARMV7 - spec = &op_armv7_spec; -#endif + oprofile_add_trace(buftail[0].lr); - if (spec) { - ret = spec->init(); - if (ret < 0) - return ret; + /* frame pointers should strictly progress back up the stack + * (towards higher addresses) */ + if (tail >= buftail[0].fp) + return NULL; - counter_config = kcalloc(spec->num_counters, sizeof(struct op_counter_config), - GFP_KERNEL); - if (!counter_config) - return -ENOMEM; + return buftail[0].fp-1; +} - op_arm_model = spec; - init_driverfs(); - ops->create_files = op_arm_create_files; - ops->setup = op_arm_setup; - ops->shutdown = op_arm_stop; - ops->start = op_arm_start; - ops->stop = op_arm_stop; - ops->cpu_type = op_arm_model->name; - printk(KERN_INFO "oprofile: using %s\n", spec->name); +static void arm_backtrace(struct pt_regs * const regs, unsigned int depth) +{ + struct frame_tail *tail = ((struct frame_tail *) regs->ARM_fp) - 1; + + if (!user_mode(regs)) { + struct stackframe frame; + frame.fp = regs->ARM_fp; + frame.sp = regs->ARM_sp; + frame.lr = regs->ARM_lr; + frame.pc = regs->ARM_pc; + walk_stackframe(&frame, report_trace, &depth); + return; } + while (depth-- && tail && !((unsigned long) tail & 3)) + tail = user_backtrace(tail); +} + +int __init oprofile_arch_init(struct oprofile_operations *ops) +{ + int cpu, ret = 0; + + perf_num_counters = perf_get_max_events(); + + counter_config = kcalloc(perf_num_counters, + sizeof(struct op_counter_config), GFP_KERNEL); + + if (!counter_config) { + pr_info("oprofile: failed to allocate %d " + "counters\n", perf_num_counters); + return -ENOMEM; + } + + for_each_possible_cpu(cpu) { + perf_events[cpu] = kcalloc(perf_num_counters, + sizeof(struct perf_event *), GFP_KERNEL); + if (!perf_events[cpu]) { + pr_info("oprofile: failed to allocate %d perf events " + "for cpu %d\n", perf_num_counters, cpu); + while (--cpu >= 0) + kfree(perf_events[cpu]); + return -ENOMEM; + } + } + + init_driverfs(); + ops->backtrace = arm_backtrace; + ops->create_files = op_arm_create_files; + ops->setup = op_arm_setup; + ops->shutdown = op_arm_stop; + ops->start = op_arm_start; + ops->stop = op_arm_stop; + ops->cpu_type = op_name_from_perf_id(armpmu_get_pmu_id()); + + if (!ops->cpu_type) + ret = -ENODEV; + else + pr_info("oprofile: using %s\n", ops->cpu_type); + return ret; } void oprofile_arch_exit(void) { - if (op_arm_model) { + int cpu, id; + struct perf_event *event; + + if (*perf_events) { exit_driverfs(); - op_arm_model = NULL; + for_each_possible_cpu(cpu) { + for (id = 0; id < perf_num_counters; ++id) { + event = perf_events[cpu][id]; + if (event != NULL) + perf_event_release_kernel(event); + } + kfree(perf_events[cpu]); + } } - kfree(counter_config); + + if (counter_config) + kfree(counter_config); +} +#else +int __init oprofile_arch_init(struct oprofile_operations *ops) +{ + pr_info("oprofile: hardware counters not available\n"); + return -ENODEV; } +void oprofile_arch_exit(void) {} +#endif /* CONFIG_HW_PERF_EVENTS */ -- 1.6.3.3 ^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 6/6] ARM: oprofile: remove old files and update KConfig 2010-03-10 10:41 ` [PATCH 5/6] ARM: oprofile: use perf-events framework as backend Will Deacon @ 2010-03-10 10:41 ` Will Deacon 0 siblings, 0 replies; 19+ messages in thread From: Will Deacon @ 2010-03-10 10:41 UTC (permalink / raw) To: linux-arm-kernel Enable hardware perf-events if CPU_HAS_PMU and select HAVE_OPROFILE if HAVE_PERF_EVENTS. If no hardware support is present, oprofile will fall back to timer mode. This patch also removes the old OProfile drivers in favour of the code implemented by perf. Signed-off-by: Will Deacon <will.deacon@arm.com> --- arch/arm/Kconfig | 26 +-- arch/arm/oprofile/Makefile | 7 +- arch/arm/oprofile/backtrace.c | 83 ------ arch/arm/oprofile/op_arm_model.h | 35 --- arch/arm/oprofile/op_counter.h | 27 -- arch/arm/oprofile/op_model_arm11_core.c | 162 ----------- arch/arm/oprofile/op_model_arm11_core.h | 45 --- arch/arm/oprofile/op_model_mpcore.c | 306 --------------------- arch/arm/oprofile/op_model_mpcore.h | 61 ----- arch/arm/oprofile/op_model_v6.c | 78 ------ arch/arm/oprofile/op_model_v7.c | 415 ----------------------------- arch/arm/oprofile/op_model_v7.h | 103 ------- arch/arm/oprofile/op_model_xscale.c | 444 ------------------------------- 13 files changed, 3 insertions(+), 1789 deletions(-) delete mode 100644 arch/arm/oprofile/backtrace.c delete mode 100644 arch/arm/oprofile/op_arm_model.h delete mode 100644 arch/arm/oprofile/op_counter.h delete mode 100644 arch/arm/oprofile/op_model_arm11_core.c delete mode 100644 arch/arm/oprofile/op_model_arm11_core.h delete mode 100644 arch/arm/oprofile/op_model_mpcore.c delete mode 100644 arch/arm/oprofile/op_model_mpcore.h delete mode 100644 arch/arm/oprofile/op_model_v6.c delete mode 100644 arch/arm/oprofile/op_model_v7.c delete mode 100644 arch/arm/oprofile/op_model_v7.h delete mode 100644 arch/arm/oprofile/op_model_xscale.c diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index df1d37c..648b610 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -12,7 +12,6 @@ config ARM select HAVE_IDE select RTC_LIB select SYS_SUPPORTS_APM_EMULATION - select HAVE_OPROFILE select HAVE_ARCH_KGDB select HAVE_KPROBES if (!XIP_KERNEL) select HAVE_KRETPROBES if (HAVE_KPROBES) @@ -23,6 +22,7 @@ config ARM select GENERIC_ATOMIC64 if (!CPU_32v6K) select HAVE_PERF_EVENTS select PERF_USE_VMALLOC + select HAVE_OPROFILE if (HAVE_PERF_EVENTS) help The ARM series is a line of low-power-consumption RISC chip designs licensed by ARM Ltd and targeted at embedded applications and @@ -164,28 +164,6 @@ config ARCH_MTD_XIP config GENERIC_HARDIRQS_NO__DO_IRQ def_bool y -if OPROFILE - -config OPROFILE_ARMV6 - def_bool y - depends on CPU_V6 && !SMP - select OPROFILE_ARM11_CORE - -config OPROFILE_MPCORE - def_bool y - depends on CPU_V6 && SMP - select OPROFILE_ARM11_CORE - -config OPROFILE_ARM11_CORE - bool - -config OPROFILE_ARMV7 - def_bool y - depends on CPU_V7 && !SMP - bool - -endif - config VECTORS_BASE hex default 0xffff0000 if MMU || CPU_HIGH_VECTOR @@ -1181,7 +1159,7 @@ config HIGHPTE config HW_PERF_EVENTS bool "Enable hardware performance counter support for perf events" - depends on PERF_EVENTS && CPU_HAS_PMU && (CPU_V6 || CPU_V7) + depends on PERF_EVENTS && CPU_HAS_PMU default y help Enable hardware performance counter support for perf events. If diff --git a/arch/arm/oprofile/Makefile b/arch/arm/oprofile/Makefile index 88e31f5..e666eaf 100644 --- a/arch/arm/oprofile/Makefile +++ b/arch/arm/oprofile/Makefile @@ -6,9 +6,4 @@ DRIVER_OBJS = $(addprefix ../../../drivers/oprofile/, \ oprofilefs.o oprofile_stats.o \ timer_int.o ) -oprofile-y := $(DRIVER_OBJS) common.o backtrace.o -oprofile-$(CONFIG_CPU_XSCALE) += op_model_xscale.o -oprofile-$(CONFIG_OPROFILE_ARM11_CORE) += op_model_arm11_core.o -oprofile-$(CONFIG_OPROFILE_ARMV6) += op_model_v6.o -oprofile-$(CONFIG_OPROFILE_MPCORE) += op_model_mpcore.o -oprofile-$(CONFIG_OPROFILE_ARMV7) += op_model_v7.o +oprofile-y := $(DRIVER_OBJS) common.o diff --git a/arch/arm/oprofile/backtrace.c b/arch/arm/oprofile/backtrace.c deleted file mode 100644 index d805a52..0000000 --- a/arch/arm/oprofile/backtrace.c +++ /dev/null @@ -1,83 +0,0 @@ -/* - * Arm specific backtracing code for oprofile - * - * Copyright 2005 Openedhand Ltd. - * - * Author: Richard Purdie <rpurdie@openedhand.com> - * - * Based on i386 oprofile backtrace code by John Levon, David Smith - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - * - */ - -#include <linux/oprofile.h> -#include <linux/sched.h> -#include <linux/mm.h> -#include <linux/uaccess.h> -#include <asm/ptrace.h> -#include <asm/stacktrace.h> - -static int report_trace(struct stackframe *frame, void *d) -{ - unsigned int *depth = d; - - if (*depth) { - oprofile_add_trace(frame->pc); - (*depth)--; - } - - return *depth == 0; -} - -/* - * The registers we're interested in are at the end of the variable - * length saved register structure. The fp points at the end of this - * structure so the address of this struct is: - * (struct frame_tail *)(xxx->fp)-1 - */ -struct frame_tail { - struct frame_tail *fp; - unsigned long sp; - unsigned long lr; -} __attribute__((packed)); - -static struct frame_tail* user_backtrace(struct frame_tail *tail) -{ - struct frame_tail buftail[2]; - - /* Also check accessibility of one struct frame_tail beyond */ - if (!access_ok(VERIFY_READ, tail, sizeof(buftail))) - return NULL; - if (__copy_from_user_inatomic(buftail, tail, sizeof(buftail))) - return NULL; - - oprofile_add_trace(buftail[0].lr); - - /* frame pointers should strictly progress back up the stack - * (towards higher addresses) */ - if (tail >= buftail[0].fp) - return NULL; - - return buftail[0].fp-1; -} - -void arm_backtrace(struct pt_regs * const regs, unsigned int depth) -{ - struct frame_tail *tail = ((struct frame_tail *) regs->ARM_fp) - 1; - - if (!user_mode(regs)) { - struct stackframe frame; - frame.fp = regs->ARM_fp; - frame.sp = regs->ARM_sp; - frame.lr = regs->ARM_lr; - frame.pc = regs->ARM_pc; - walk_stackframe(&frame, report_trace, &depth); - return; - } - - while (depth-- && tail && !((unsigned long) tail & 3)) - tail = user_backtrace(tail); -} diff --git a/arch/arm/oprofile/op_arm_model.h b/arch/arm/oprofile/op_arm_model.h deleted file mode 100644 index 8c4e4f6..0000000 --- a/arch/arm/oprofile/op_arm_model.h +++ /dev/null @@ -1,35 +0,0 @@ -/** - * @file op_arm_model.h - * interface to ARM machine specific operations - * - * @remark Copyright 2004 Oprofile Authors - * @remark Read the file COPYING - * - * @author Zwane Mwaikambo - */ - -#ifndef OP_ARM_MODEL_H -#define OP_ARM_MODEL_H - -struct op_arm_model_spec { - int (*init)(void); - unsigned int num_counters; - int (*setup_ctrs)(void); - int (*start)(void); - void (*stop)(void); - char *name; -}; - -#ifdef CONFIG_CPU_XSCALE -extern struct op_arm_model_spec op_xscale_spec; -#endif - -extern struct op_arm_model_spec op_armv6_spec; -extern struct op_arm_model_spec op_mpcore_spec; -extern struct op_arm_model_spec op_armv7_spec; - -extern void arm_backtrace(struct pt_regs * const regs, unsigned int depth); - -extern int __init op_arm_init(struct oprofile_operations *ops, struct op_arm_model_spec *spec); -extern void op_arm_exit(void); -#endif /* OP_ARM_MODEL_H */ diff --git a/arch/arm/oprofile/op_counter.h b/arch/arm/oprofile/op_counter.h deleted file mode 100644 index ca942a6..0000000 --- a/arch/arm/oprofile/op_counter.h +++ /dev/null @@ -1,27 +0,0 @@ -/** - * @file op_counter.h - * - * @remark Copyright 2004 Oprofile Authors - * @remark Read the file COPYING - * - * @author Zwane Mwaikambo - */ - -#ifndef OP_COUNTER_H -#define OP_COUNTER_H - -/* Per performance monitor configuration as set via - * oprofilefs. - */ -struct op_counter_config { - unsigned long count; - unsigned long enabled; - unsigned long event; - unsigned long unit_mask; - unsigned long kernel; - unsigned long user; -}; - -extern struct op_counter_config *counter_config; - -#endif /* OP_COUNTER_H */ diff --git a/arch/arm/oprofile/op_model_arm11_core.c b/arch/arm/oprofile/op_model_arm11_core.c deleted file mode 100644 index ef3e265..0000000 --- a/arch/arm/oprofile/op_model_arm11_core.c +++ /dev/null @@ -1,162 +0,0 @@ -/** - * @file op_model_arm11_core.c - * ARM11 Event Monitor Driver - * @remark Copyright 2004 ARM SMP Development Team - */ -#include <linux/types.h> -#include <linux/errno.h> -#include <linux/oprofile.h> -#include <linux/interrupt.h> -#include <linux/irq.h> -#include <linux/smp.h> - -#include "op_counter.h" -#include "op_arm_model.h" -#include "op_model_arm11_core.h" - -/* - * ARM11 PMU support - */ -static inline void arm11_write_pmnc(u32 val) -{ - /* upper 4bits and 7, 11 are write-as-0 */ - val &= 0x0ffff77f; - asm volatile("mcr p15, 0, %0, c15, c12, 0" : : "r" (val)); -} - -static inline u32 arm11_read_pmnc(void) -{ - u32 val; - asm volatile("mrc p15, 0, %0, c15, c12, 0" : "=r" (val)); - return val; -} - -static void arm11_reset_counter(unsigned int cnt) -{ - u32 val = -(u32)counter_config[CPU_COUNTER(smp_processor_id(), cnt)].count; - switch (cnt) { - case CCNT: - asm volatile("mcr p15, 0, %0, c15, c12, 1" : : "r" (val)); - break; - - case PMN0: - asm volatile("mcr p15, 0, %0, c15, c12, 2" : : "r" (val)); - break; - - case PMN1: - asm volatile("mcr p15, 0, %0, c15, c12, 3" : : "r" (val)); - break; - } -} - -int arm11_setup_pmu(void) -{ - unsigned int cnt; - u32 pmnc; - - if (arm11_read_pmnc() & PMCR_E) { - printk(KERN_ERR "oprofile: CPU%u PMU still enabled when setup new event counter.\n", smp_processor_id()); - return -EBUSY; - } - - /* initialize PMNC, reset overflow, D bit, C bit and P bit. */ - arm11_write_pmnc(PMCR_OFL_PMN0 | PMCR_OFL_PMN1 | PMCR_OFL_CCNT | - PMCR_C | PMCR_P); - - for (pmnc = 0, cnt = PMN0; cnt <= CCNT; cnt++) { - unsigned long event; - - if (!counter_config[CPU_COUNTER(smp_processor_id(), cnt)].enabled) - continue; - - event = counter_config[CPU_COUNTER(smp_processor_id(), cnt)].event & 255; - - /* - * Set event (if destined for PMNx counters) - */ - if (cnt == PMN0) { - pmnc |= event << 20; - } else if (cnt == PMN1) { - pmnc |= event << 12; - } - - /* - * We don't need to set the event if it's a cycle count - * Enable interrupt for this counter - */ - pmnc |= PMCR_IEN_PMN0 << cnt; - arm11_reset_counter(cnt); - } - arm11_write_pmnc(pmnc); - - return 0; -} - -int arm11_start_pmu(void) -{ - arm11_write_pmnc(arm11_read_pmnc() | PMCR_E); - return 0; -} - -int arm11_stop_pmu(void) -{ - unsigned int cnt; - - arm11_write_pmnc(arm11_read_pmnc() & ~PMCR_E); - - for (cnt = PMN0; cnt <= CCNT; cnt++) - arm11_reset_counter(cnt); - - return 0; -} - -/* - * CPU counters' IRQ handler (one IRQ per CPU) - */ -static irqreturn_t arm11_pmu_interrupt(int irq, void *arg) -{ - struct pt_regs *regs = get_irq_regs(); - unsigned int cnt; - u32 pmnc; - - pmnc = arm11_read_pmnc(); - - for (cnt = PMN0; cnt <= CCNT; cnt++) { - if ((pmnc & (PMCR_OFL_PMN0 << cnt)) && (pmnc & (PMCR_IEN_PMN0 << cnt))) { - arm11_reset_counter(cnt); - oprofile_add_sample(regs, CPU_COUNTER(smp_processor_id(), cnt)); - } - } - /* Clear counter flag(s) */ - arm11_write_pmnc(pmnc); - return IRQ_HANDLED; -} - -int arm11_request_interrupts(const int *irqs, int nr) -{ - unsigned int i; - int ret = 0; - - for(i = 0; i < nr; i++) { - ret = request_irq(irqs[i], arm11_pmu_interrupt, IRQF_DISABLED, "CP15 PMU", NULL); - if (ret != 0) { - printk(KERN_ERR "oprofile: unable to request IRQ%u for MPCORE-EM\n", - irqs[i]); - break; - } - } - - if (i != nr) - while (i-- != 0) - free_irq(irqs[i], NULL); - - return ret; -} - -void arm11_release_interrupts(const int *irqs, int nr) -{ - unsigned int i; - - for (i = 0; i < nr; i++) - free_irq(irqs[i], NULL); -} diff --git a/arch/arm/oprofile/op_model_arm11_core.h b/arch/arm/oprofile/op_model_arm11_core.h deleted file mode 100644 index 1902b99..0000000 --- a/arch/arm/oprofile/op_model_arm11_core.h +++ /dev/null @@ -1,45 +0,0 @@ -/** - * @file op_model_arm11_core.h - * ARM11 Event Monitor Driver - * @remark Copyright 2004 ARM SMP Development Team - * @remark Copyright 2000-2004 Deepak Saxena <dsaxena@mvista.com> - * @remark Copyright 2000-2004 MontaVista Software Inc - * @remark Copyright 2004 Dave Jiang <dave.jiang@intel.com> - * @remark Copyright 2004 Intel Corporation - * @remark Copyright 2004 Zwane Mwaikambo <zwane@arm.linux.org.uk> - * @remark Copyright 2004 Oprofile Authors - * - * @remark Read the file COPYING - * - * @author Zwane Mwaikambo - */ -#ifndef OP_MODEL_ARM11_CORE_H -#define OP_MODEL_ARM11_CORE_H - -/* - * Per-CPU PMCR - */ -#define PMCR_E (1 << 0) /* Enable */ -#define PMCR_P (1 << 1) /* Count reset */ -#define PMCR_C (1 << 2) /* Cycle counter reset */ -#define PMCR_D (1 << 3) /* Cycle counter counts every 64th cpu cycle */ -#define PMCR_IEN_PMN0 (1 << 4) /* Interrupt enable count reg 0 */ -#define PMCR_IEN_PMN1 (1 << 5) /* Interrupt enable count reg 1 */ -#define PMCR_IEN_CCNT (1 << 6) /* Interrupt enable cycle counter */ -#define PMCR_OFL_PMN0 (1 << 8) /* Count reg 0 overflow */ -#define PMCR_OFL_PMN1 (1 << 9) /* Count reg 1 overflow */ -#define PMCR_OFL_CCNT (1 << 10) /* Cycle counter overflow */ - -#define PMN0 0 -#define PMN1 1 -#define CCNT 2 - -#define CPU_COUNTER(cpu, counter) ((cpu) * 3 + (counter)) - -int arm11_setup_pmu(void); -int arm11_start_pmu(void); -int arm11_stop_pmu(void); -int arm11_request_interrupts(const int *, int); -void arm11_release_interrupts(const int *, int); - -#endif diff --git a/arch/arm/oprofile/op_model_mpcore.c b/arch/arm/oprofile/op_model_mpcore.c deleted file mode 100644 index f73ce87..0000000 --- a/arch/arm/oprofile/op_model_mpcore.c +++ /dev/null @@ -1,306 +0,0 @@ -/** - * @file op_model_mpcore.c - * MPCORE Event Monitor Driver - * @remark Copyright 2004 ARM SMP Development Team - * @remark Copyright 2000-2004 Deepak Saxena <dsaxena@mvista.com> - * @remark Copyright 2000-2004 MontaVista Software Inc - * @remark Copyright 2004 Dave Jiang <dave.jiang@intel.com> - * @remark Copyright 2004 Intel Corporation - * @remark Copyright 2004 Zwane Mwaikambo <zwane@arm.linux.org.uk> - * @remark Copyright 2004 Oprofile Authors - * - * @remark Read the file COPYING - * - * @author Zwane Mwaikambo - * - * Counters: - * 0: PMN0 on CPU0, per-cpu configurable event counter - * 1: PMN1 on CPU0, per-cpu configurable event counter - * 2: CCNT on CPU0 - * 3: PMN0 on CPU1 - * 4: PMN1 on CPU1 - * 5: CCNT on CPU1 - * 6: PMN0 on CPU1 - * 7: PMN1 on CPU1 - * 8: CCNT on CPU1 - * 9: PMN0 on CPU1 - * 10: PMN1 on CPU1 - * 11: CCNT on CPU1 - * 12-19: configurable SCU event counters - */ - -/* #define DEBUG */ -#include <linux/types.h> -#include <linux/errno.h> -#include <linux/err.h> -#include <linux/sched.h> -#include <linux/oprofile.h> -#include <linux/interrupt.h> -#include <linux/smp.h> -#include <linux/io.h> - -#include <asm/irq.h> -#include <asm/mach/irq.h> -#include <mach/hardware.h> -#include <mach/board-eb.h> -#include <asm/system.h> -#include <asm/pmu.h> - -#include "op_counter.h" -#include "op_arm_model.h" -#include "op_model_arm11_core.h" -#include "op_model_mpcore.h" - -/* - * MPCore SCU event monitor support - */ -#define SCU_EVENTMONITORS_VA_BASE __io_address(REALVIEW_EB11MP_SCU_BASE + 0x10) - -/* - * Bitmask of used SCU counters - */ -static unsigned int scu_em_used; -static const struct pmu_irqs *pmu_irqs; - -/* - * 2 helper fns take a counter number from 0-7 (not the userspace-visible counter number) - */ -static inline void scu_reset_counter(struct eventmonitor __iomem *emc, unsigned int n) -{ - writel(-(u32)counter_config[SCU_COUNTER(n)].count, &emc->MC[n]); -} - -static inline void scu_set_event(struct eventmonitor __iomem *emc, unsigned int n, u32 event) -{ - event &= 0xff; - writeb(event, &emc->MCEB[n]); -} - -/* - * SCU counters' IRQ handler (one IRQ per counter => 2 IRQs per CPU) - */ -static irqreturn_t scu_em_interrupt(int irq, void *arg) -{ - struct eventmonitor __iomem *emc = SCU_EVENTMONITORS_VA_BASE; - unsigned int cnt; - - cnt = irq - IRQ_EB11MP_PMU_SCU0; - oprofile_add_sample(get_irq_regs(), SCU_COUNTER(cnt)); - scu_reset_counter(emc, cnt); - - /* Clear overflow flag for this counter */ - writel(1 << (cnt + 16), &emc->PMCR); - - return IRQ_HANDLED; -} - -/* Configure just the SCU counters that the user has requested */ -static void scu_setup(void) -{ - struct eventmonitor __iomem *emc = SCU_EVENTMONITORS_VA_BASE; - unsigned int i; - - scu_em_used = 0; - - for (i = 0; i < NUM_SCU_COUNTERS; i++) { - if (counter_config[SCU_COUNTER(i)].enabled && - counter_config[SCU_COUNTER(i)].event) { - scu_set_event(emc, i, 0); /* disable counter for now */ - scu_em_used |= 1 << i; - } - } -} - -static int scu_start(void) -{ - struct eventmonitor __iomem *emc = SCU_EVENTMONITORS_VA_BASE; - unsigned int temp, i; - unsigned long event; - int ret = 0; - - /* - * request the SCU counter interrupts that we need - */ - for (i = 0; i < NUM_SCU_COUNTERS; i++) { - if (scu_em_used & (1 << i)) { - ret = request_irq(IRQ_EB11MP_PMU_SCU0 + i, scu_em_interrupt, IRQF_DISABLED, "SCU PMU", NULL); - if (ret) { - printk(KERN_ERR "oprofile: unable to request IRQ%u for SCU Event Monitor\n", - IRQ_EB11MP_PMU_SCU0 + i); - goto err_free_scu; - } - } - } - - /* - * clear overflow and enable interrupt for all used counters - */ - temp = readl(&emc->PMCR); - for (i = 0; i < NUM_SCU_COUNTERS; i++) { - if (scu_em_used & (1 << i)) { - scu_reset_counter(emc, i); - event = counter_config[SCU_COUNTER(i)].event; - scu_set_event(emc, i, event); - - /* clear overflow/interrupt */ - temp |= 1 << (i + 16); - /* enable interrupt*/ - temp |= 1 << (i + 8); - } - } - - /* Enable all 8 counters */ - temp |= PMCR_E; - writel(temp, &emc->PMCR); - - return 0; - - err_free_scu: - while (i--) - free_irq(IRQ_EB11MP_PMU_SCU0 + i, NULL); - return ret; -} - -static void scu_stop(void) -{ - struct eventmonitor __iomem *emc = SCU_EVENTMONITORS_VA_BASE; - unsigned int temp, i; - - /* Disable counter interrupts */ - /* Don't disable all 8 counters (with the E bit) as they may be in use */ - temp = readl(&emc->PMCR); - for (i = 0; i < NUM_SCU_COUNTERS; i++) { - if (scu_em_used & (1 << i)) - temp &= ~(1 << (i + 8)); - } - writel(temp, &emc->PMCR); - - /* Free counter interrupts and reset counters */ - for (i = 0; i < NUM_SCU_COUNTERS; i++) { - if (scu_em_used & (1 << i)) { - scu_reset_counter(emc, i); - free_irq(IRQ_EB11MP_PMU_SCU0 + i, NULL); - } - } -} - -struct em_function_data { - int (*fn)(void); - int ret; -}; - -static void em_func(void *data) -{ - struct em_function_data *d = data; - int ret = d->fn(); - if (ret) - d->ret = ret; -} - -static int em_call_function(int (*fn)(void)) -{ - struct em_function_data data; - - data.fn = fn; - data.ret = 0; - - preempt_disable(); - smp_call_function(em_func, &data, 1); - em_func(&data); - preempt_enable(); - - return data.ret; -} - -/* - * Glue to stick the individual ARM11 PMUs and the SCU - * into the oprofile framework. - */ -static int em_setup_ctrs(void) -{ - int ret; - - /* Configure CPU counters by cross-calling to the other CPUs */ - ret = em_call_function(arm11_setup_pmu); - if (ret == 0) - scu_setup(); - - return 0; -} - -static int em_start(void) -{ - int ret; - - pmu_irqs = reserve_pmu(); - if (IS_ERR(pmu_irqs)) { - ret = PTR_ERR(pmu_irqs); - goto out; - } - - ret = arm11_request_interrupts(pmu_irqs->irqs, pmu_irqs->num_irqs); - if (ret == 0) { - em_call_function(arm11_start_pmu); - - ret = scu_start(); - if (ret) { - arm11_release_interrupts(pmu_irqs->irqs, - pmu_irqs->num_irqs); - } else { - release_pmu(pmu_irqs); - pmu_irqs = NULL; - } - } - -out: - return ret; -} - -static void em_stop(void) -{ - em_call_function(arm11_stop_pmu); - arm11_release_interrupts(pmu_irqs->irqs, pmu_irqs->num_irqs); - scu_stop(); - release_pmu(pmu_irqs); -} - -/* - * Why isn't there a function to route an IRQ to a specific CPU in - * genirq? - */ -static void em_route_irq(int irq, unsigned int cpu) -{ - struct irq_desc *desc = irq_desc + irq; - const struct cpumask *mask = cpumask_of(cpu); - - spin_lock_irq(&desc->lock); - cpumask_copy(desc->affinity, mask); - desc->chip->set_affinity(irq, mask); - spin_unlock_irq(&desc->lock); -} - -static int em_setup(void) -{ - /* - * Send SCU PMU interrupts to the "owner" CPU. - */ - em_route_irq(IRQ_EB11MP_PMU_SCU0, 0); - em_route_irq(IRQ_EB11MP_PMU_SCU1, 0); - em_route_irq(IRQ_EB11MP_PMU_SCU2, 1); - em_route_irq(IRQ_EB11MP_PMU_SCU3, 1); - em_route_irq(IRQ_EB11MP_PMU_SCU4, 2); - em_route_irq(IRQ_EB11MP_PMU_SCU5, 2); - em_route_irq(IRQ_EB11MP_PMU_SCU6, 3); - em_route_irq(IRQ_EB11MP_PMU_SCU7, 3); - - return init_pmu(); -} - -struct op_arm_model_spec op_mpcore_spec = { - .init = em_setup, - .num_counters = MPCORE_NUM_COUNTERS, - .setup_ctrs = em_setup_ctrs, - .start = em_start, - .stop = em_stop, - .name = "arm/mpcore", -}; diff --git a/arch/arm/oprofile/op_model_mpcore.h b/arch/arm/oprofile/op_model_mpcore.h deleted file mode 100644 index 73d8110..0000000 --- a/arch/arm/oprofile/op_model_mpcore.h +++ /dev/null @@ -1,61 +0,0 @@ -/** - * @file op_model_mpcore.c - * MPCORE Event Monitor Driver - * @remark Copyright 2004 ARM SMP Development Team - * @remark Copyright 2000-2004 Deepak Saxena <dsaxena@mvista.com> - * @remark Copyright 2000-2004 MontaVista Software Inc - * @remark Copyright 2004 Dave Jiang <dave.jiang@intel.com> - * @remark Copyright 2004 Intel Corporation - * @remark Copyright 2004 Zwane Mwaikambo <zwane@arm.linux.org.uk> - * @remark Copyright 2004 Oprofile Authors - * - * @remark Read the file COPYING - * - * @author Zwane Mwaikambo - */ -#ifndef OP_MODEL_MPCORE_H -#define OP_MODEL_MPCORE_H - -struct eventmonitor { - unsigned long PMCR; - unsigned char MCEB[8]; - unsigned long MC[8]; -}; - -/* - * List of userspace counter numbers: note that the structure is important. - * The code relies on CPUn's counters being CPU0's counters + 3n - * and on CPU0's counters starting at 0 - */ - -#define COUNTER_CPU0_PMN0 0 -#define COUNTER_CPU0_PMN1 1 -#define COUNTER_CPU0_CCNT 2 - -#define COUNTER_CPU1_PMN0 3 -#define COUNTER_CPU1_PMN1 4 -#define COUNTER_CPU1_CCNT 5 - -#define COUNTER_CPU2_PMN0 6 -#define COUNTER_CPU2_PMN1 7 -#define COUNTER_CPU2_CCNT 8 - -#define COUNTER_CPU3_PMN0 9 -#define COUNTER_CPU3_PMN1 10 -#define COUNTER_CPU3_CCNT 11 - -#define COUNTER_SCU_MN0 12 -#define COUNTER_SCU_MN1 13 -#define COUNTER_SCU_MN2 14 -#define COUNTER_SCU_MN3 15 -#define COUNTER_SCU_MN4 16 -#define COUNTER_SCU_MN5 17 -#define COUNTER_SCU_MN6 18 -#define COUNTER_SCU_MN7 19 -#define NUM_SCU_COUNTERS 8 - -#define SCU_COUNTER(number) ((number) + COUNTER_SCU_MN0) - -#define MPCORE_NUM_COUNTERS SCU_COUNTER(NUM_SCU_COUNTERS) - -#endif diff --git a/arch/arm/oprofile/op_model_v6.c b/arch/arm/oprofile/op_model_v6.c deleted file mode 100644 index a22357a..0000000 --- a/arch/arm/oprofile/op_model_v6.c +++ /dev/null @@ -1,78 +0,0 @@ -/** - * @file op_model_v6.c - * ARM11 Performance Monitor Driver - * - * Based on op_model_xscale.c - * - * @remark Copyright 2000-2004 Deepak Saxena <dsaxena@mvista.com> - * @remark Copyright 2000-2004 MontaVista Software Inc - * @remark Copyright 2004 Dave Jiang <dave.jiang@intel.com> - * @remark Copyright 2004 Intel Corporation - * @remark Copyright 2004 Zwane Mwaikambo <zwane@arm.linux.org.uk> - * @remark Copyright 2004 OProfile Authors - * - * @remark Read the file COPYING - * - * @author Tony Lindgren <tony@atomide.com> - */ - -/* #define DEBUG */ -#include <linux/types.h> -#include <linux/errno.h> -#include <linux/err.h> -#include <linux/sched.h> -#include <linux/oprofile.h> -#include <linux/interrupt.h> -#include <asm/irq.h> -#include <asm/system.h> -#include <asm/pmu.h> - -#include "op_counter.h" -#include "op_arm_model.h" -#include "op_model_arm11_core.h" - -static const struct pmu_irqs *pmu_irqs; - -static void armv6_pmu_stop(void) -{ - arm11_stop_pmu(); - arm11_release_interrupts(pmu_irqs->irqs, pmu_irqs->num_irqs); - release_pmu(pmu_irqs); - pmu_irqs = NULL; -} - -static int armv6_pmu_start(void) -{ - int ret; - - pmu_irqs = reserve_pmu(); - if (IS_ERR(pmu_irqs)) { - ret = PTR_ERR(pmu_irqs); - goto out; - } - - ret = arm11_request_interrupts(pmu_irqs->irqs, pmu_irqs->num_irqs); - if (ret >= 0) { - ret = arm11_start_pmu(); - } else { - release_pmu(pmu_irqs); - pmu_irqs = NULL; - } - -out: - return ret; -} - -static int armv6_detect_pmu(void) -{ - return 0; -} - -struct op_arm_model_spec op_armv6_spec = { - .init = armv6_detect_pmu, - .num_counters = 3, - .setup_ctrs = arm11_setup_pmu, - .start = armv6_pmu_start, - .stop = armv6_pmu_stop, - .name = "arm/armv6", -}; diff --git a/arch/arm/oprofile/op_model_v7.c b/arch/arm/oprofile/op_model_v7.c deleted file mode 100644 index 8642d08..0000000 --- a/arch/arm/oprofile/op_model_v7.c +++ /dev/null @@ -1,415 +0,0 @@ -/** - * op_model_v7.c - * ARM V7 (Cortex A8) Event Monitor Driver - * - * Copyright 2008 Jean Pihet <jpihet@mvista.com> - * Copyright 2004 ARM SMP Development Team - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - */ -#include <linux/types.h> -#include <linux/errno.h> -#include <linux/err.h> -#include <linux/oprofile.h> -#include <linux/interrupt.h> -#include <linux/irq.h> -#include <linux/smp.h> - -#include <asm/pmu.h> - -#include "op_counter.h" -#include "op_arm_model.h" -#include "op_model_v7.h" - -/* #define DEBUG */ - - -/* - * ARM V7 PMNC support - */ - -static u32 cnt_en[CNTMAX]; - -static inline void armv7_pmnc_write(u32 val) -{ - val &= PMNC_MASK; - asm volatile("mcr p15, 0, %0, c9, c12, 0" : : "r" (val)); -} - -static inline u32 armv7_pmnc_read(void) -{ - u32 val; - - asm volatile("mrc p15, 0, %0, c9, c12, 0" : "=r" (val)); - return val; -} - -static inline u32 armv7_pmnc_enable_counter(unsigned int cnt) -{ - u32 val; - - if (cnt >= CNTMAX) { - printk(KERN_ERR "oprofile: CPU%u enabling wrong PMNC counter" - " %d\n", smp_processor_id(), cnt); - return -1; - } - - if (cnt == CCNT) - val = CNTENS_C; - else - val = (1 << (cnt - CNT0)); - - val &= CNTENS_MASK; - asm volatile("mcr p15, 0, %0, c9, c12, 1" : : "r" (val)); - - return cnt; -} - -static inline u32 armv7_pmnc_disable_counter(unsigned int cnt) -{ - u32 val; - - if (cnt >= CNTMAX) { - printk(KERN_ERR "oprofile: CPU%u disabling wrong PMNC counter" - " %d\n", smp_processor_id(), cnt); - return -1; - } - - if (cnt == CCNT) - val = CNTENC_C; - else - val = (1 << (cnt - CNT0)); - - val &= CNTENC_MASK; - asm volatile("mcr p15, 0, %0, c9, c12, 2" : : "r" (val)); - - return cnt; -} - -static inline u32 armv7_pmnc_enable_intens(unsigned int cnt) -{ - u32 val; - - if (cnt >= CNTMAX) { - printk(KERN_ERR "oprofile: CPU%u enabling wrong PMNC counter" - " interrupt enable %d\n", smp_processor_id(), cnt); - return -1; - } - - if (cnt == CCNT) - val = INTENS_C; - else - val = (1 << (cnt - CNT0)); - - val &= INTENS_MASK; - asm volatile("mcr p15, 0, %0, c9, c14, 1" : : "r" (val)); - - return cnt; -} - -static inline u32 armv7_pmnc_getreset_flags(void) -{ - u32 val; - - /* Read */ - asm volatile("mrc p15, 0, %0, c9, c12, 3" : "=r" (val)); - - /* Write to clear flags */ - val &= FLAG_MASK; - asm volatile("mcr p15, 0, %0, c9, c12, 3" : : "r" (val)); - - return val; -} - -static inline int armv7_pmnc_select_counter(unsigned int cnt) -{ - u32 val; - - if ((cnt == CCNT) || (cnt >= CNTMAX)) { - printk(KERN_ERR "oprofile: CPU%u selecting wrong PMNC counteri" - " %d\n", smp_processor_id(), cnt); - return -1; - } - - val = (cnt - CNT0) & SELECT_MASK; - asm volatile("mcr p15, 0, %0, c9, c12, 5" : : "r" (val)); - - return cnt; -} - -static inline void armv7_pmnc_write_evtsel(unsigned int cnt, u32 val) -{ - if (armv7_pmnc_select_counter(cnt) == cnt) { - val &= EVTSEL_MASK; - asm volatile("mcr p15, 0, %0, c9, c13, 1" : : "r" (val)); - } -} - -static void armv7_pmnc_reset_counter(unsigned int cnt) -{ - u32 cpu_cnt = CPU_COUNTER(smp_processor_id(), cnt); - u32 val = -(u32)counter_config[cpu_cnt].count; - - switch (cnt) { - case CCNT: - armv7_pmnc_disable_counter(cnt); - - asm volatile("mcr p15, 0, %0, c9, c13, 0" : : "r" (val)); - - if (cnt_en[cnt] != 0) - armv7_pmnc_enable_counter(cnt); - - break; - - case CNT0: - case CNT1: - case CNT2: - case CNT3: - armv7_pmnc_disable_counter(cnt); - - if (armv7_pmnc_select_counter(cnt) == cnt) - asm volatile("mcr p15, 0, %0, c9, c13, 2" : : "r" (val)); - - if (cnt_en[cnt] != 0) - armv7_pmnc_enable_counter(cnt); - - break; - - default: - printk(KERN_ERR "oprofile: CPU%u resetting wrong PMNC counter" - " %d\n", smp_processor_id(), cnt); - break; - } -} - -int armv7_setup_pmnc(void) -{ - unsigned int cnt; - - if (armv7_pmnc_read() & PMNC_E) { - printk(KERN_ERR "oprofile: CPU%u PMNC still enabled when setup" - " new event counter.\n", smp_processor_id()); - return -EBUSY; - } - - /* Initialize & Reset PMNC: C bit and P bit */ - armv7_pmnc_write(PMNC_P | PMNC_C); - - - for (cnt = CCNT; cnt < CNTMAX; cnt++) { - unsigned long event; - u32 cpu_cnt = CPU_COUNTER(smp_processor_id(), cnt); - - /* - * Disable counter - */ - armv7_pmnc_disable_counter(cnt); - cnt_en[cnt] = 0; - - if (!counter_config[cpu_cnt].enabled) - continue; - - event = counter_config[cpu_cnt].event & 255; - - /* - * Set event (if destined for PMNx counters) - * We don't need to set the event if it's a cycle count - */ - if (cnt != CCNT) - armv7_pmnc_write_evtsel(cnt, event); - - /* - * Enable interrupt for this counter - */ - armv7_pmnc_enable_intens(cnt); - - /* - * Reset counter - */ - armv7_pmnc_reset_counter(cnt); - - /* - * Enable counter - */ - armv7_pmnc_enable_counter(cnt); - cnt_en[cnt] = 1; - } - - return 0; -} - -static inline void armv7_start_pmnc(void) -{ - armv7_pmnc_write(armv7_pmnc_read() | PMNC_E); -} - -static inline void armv7_stop_pmnc(void) -{ - armv7_pmnc_write(armv7_pmnc_read() & ~PMNC_E); -} - -/* - * CPU counters' IRQ handler (one IRQ per CPU) - */ -static irqreturn_t armv7_pmnc_interrupt(int irq, void *arg) -{ - struct pt_regs *regs = get_irq_regs(); - unsigned int cnt; - u32 flags; - - - /* - * Stop IRQ generation - */ - armv7_stop_pmnc(); - - /* - * Get and reset overflow status flags - */ - flags = armv7_pmnc_getreset_flags(); - - /* - * Cycle counter - */ - if (flags & FLAG_C) { - u32 cpu_cnt = CPU_COUNTER(smp_processor_id(), CCNT); - armv7_pmnc_reset_counter(CCNT); - oprofile_add_sample(regs, cpu_cnt); - } - - /* - * PMNC counters 0:3 - */ - for (cnt = CNT0; cnt < CNTMAX; cnt++) { - if (flags & (1 << (cnt - CNT0))) { - u32 cpu_cnt = CPU_COUNTER(smp_processor_id(), cnt); - armv7_pmnc_reset_counter(cnt); - oprofile_add_sample(regs, cpu_cnt); - } - } - - /* - * Allow IRQ generation - */ - armv7_start_pmnc(); - - return IRQ_HANDLED; -} - -int armv7_request_interrupts(const int *irqs, int nr) -{ - unsigned int i; - int ret = 0; - - for (i = 0; i < nr; i++) { - ret = request_irq(irqs[i], armv7_pmnc_interrupt, - IRQF_DISABLED, "CP15 PMNC", NULL); - if (ret != 0) { - printk(KERN_ERR "oprofile: unable to request IRQ%u" - " for ARMv7\n", - irqs[i]); - break; - } - } - - if (i != nr) - while (i-- != 0) - free_irq(irqs[i], NULL); - - return ret; -} - -void armv7_release_interrupts(const int *irqs, int nr) -{ - unsigned int i; - - for (i = 0; i < nr; i++) - free_irq(irqs[i], NULL); -} - -#ifdef DEBUG -static void armv7_pmnc_dump_regs(void) -{ - u32 val; - unsigned int cnt; - - printk(KERN_INFO "PMNC registers dump:\n"); - - asm volatile("mrc p15, 0, %0, c9, c12, 0" : "=r" (val)); - printk(KERN_INFO "PMNC =0x%08x\n", val); - - asm volatile("mrc p15, 0, %0, c9, c12, 1" : "=r" (val)); - printk(KERN_INFO "CNTENS=0x%08x\n", val); - - asm volatile("mrc p15, 0, %0, c9, c14, 1" : "=r" (val)); - printk(KERN_INFO "INTENS=0x%08x\n", val); - - asm volatile("mrc p15, 0, %0, c9, c12, 3" : "=r" (val)); - printk(KERN_INFO "FLAGS =0x%08x\n", val); - - asm volatile("mrc p15, 0, %0, c9, c12, 5" : "=r" (val)); - printk(KERN_INFO "SELECT=0x%08x\n", val); - - asm volatile("mrc p15, 0, %0, c9, c13, 0" : "=r" (val)); - printk(KERN_INFO "CCNT =0x%08x\n", val); - - for (cnt = CNT0; cnt < CNTMAX; cnt++) { - armv7_pmnc_select_counter(cnt); - asm volatile("mrc p15, 0, %0, c9, c13, 2" : "=r" (val)); - printk(KERN_INFO "CNT[%d] count =0x%08x\n", cnt-CNT0, val); - asm volatile("mrc p15, 0, %0, c9, c13, 1" : "=r" (val)); - printk(KERN_INFO "CNT[%d] evtsel=0x%08x\n", cnt-CNT0, val); - } -} -#endif - -static const struct pmu_irqs *pmu_irqs; - -static void armv7_pmnc_stop(void) -{ -#ifdef DEBUG - armv7_pmnc_dump_regs(); -#endif - armv7_stop_pmnc(); - armv7_release_interrupts(pmu_irqs->irqs, pmu_irqs->num_irqs); - release_pmu(pmu_irqs); - pmu_irqs = NULL; -} - -static int armv7_pmnc_start(void) -{ - int ret; - - pmu_irqs = reserve_pmu(); - if (IS_ERR(pmu_irqs)) - return PTR_ERR(pmu_irqs); - -#ifdef DEBUG - armv7_pmnc_dump_regs(); -#endif - ret = armv7_request_interrupts(pmu_irqs->irqs, pmu_irqs->num_irqs); - if (ret >= 0) { - armv7_start_pmnc(); - } else { - release_pmu(pmu_irqs); - pmu_irqs = NULL; - } - - return ret; -} - -static int armv7_detect_pmnc(void) -{ - return 0; -} - -struct op_arm_model_spec op_armv7_spec = { - .init = armv7_detect_pmnc, - .num_counters = 5, - .setup_ctrs = armv7_setup_pmnc, - .start = armv7_pmnc_start, - .stop = armv7_pmnc_stop, - .name = "arm/armv7", -}; diff --git a/arch/arm/oprofile/op_model_v7.h b/arch/arm/oprofile/op_model_v7.h deleted file mode 100644 index 9ca334b..0000000 --- a/arch/arm/oprofile/op_model_v7.h +++ /dev/null @@ -1,103 +0,0 @@ -/** - * op_model_v7.h - * ARM v7 (Cortex A8) Event Monitor Driver - * - * Copyright 2008 Jean Pihet <jpihet@mvista.com> - * Copyright 2004 ARM SMP Development Team - * Copyright 2000-2004 Deepak Saxena <dsaxena@mvista.com> - * Copyright 2000-2004 MontaVista Software Inc - * Copyright 2004 Dave Jiang <dave.jiang@intel.com> - * Copyright 2004 Intel Corporation - * Copyright 2004 Zwane Mwaikambo <zwane@arm.linux.org.uk> - * Copyright 2004 Oprofile Authors - * - * Read the file COPYING - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - */ -#ifndef OP_MODEL_V7_H -#define OP_MODEL_V7_H - -/* - * Per-CPU PMNC: config reg - */ -#define PMNC_E (1 << 0) /* Enable all counters */ -#define PMNC_P (1 << 1) /* Reset all counters */ -#define PMNC_C (1 << 2) /* Cycle counter reset */ -#define PMNC_D (1 << 3) /* CCNT counts every 64th cpu cycle */ -#define PMNC_X (1 << 4) /* Export to ETM */ -#define PMNC_DP (1 << 5) /* Disable CCNT if non-invasive debug*/ -#define PMNC_MASK 0x3f /* Mask for writable bits */ - -/* - * Available counters - */ -#define CCNT 0 -#define CNT0 1 -#define CNT1 2 -#define CNT2 3 -#define CNT3 4 -#define CNTMAX 5 - -#define CPU_COUNTER(cpu, counter) ((cpu) * CNTMAX + (counter)) - -/* - * CNTENS: counters enable reg - */ -#define CNTENS_P0 (1 << 0) -#define CNTENS_P1 (1 << 1) -#define CNTENS_P2 (1 << 2) -#define CNTENS_P3 (1 << 3) -#define CNTENS_C (1 << 31) -#define CNTENS_MASK 0x8000000f /* Mask for writable bits */ - -/* - * CNTENC: counters disable reg - */ -#define CNTENC_P0 (1 << 0) -#define CNTENC_P1 (1 << 1) -#define CNTENC_P2 (1 << 2) -#define CNTENC_P3 (1 << 3) -#define CNTENC_C (1 << 31) -#define CNTENC_MASK 0x8000000f /* Mask for writable bits */ - -/* - * INTENS: counters overflow interrupt enable reg - */ -#define INTENS_P0 (1 << 0) -#define INTENS_P1 (1 << 1) -#define INTENS_P2 (1 << 2) -#define INTENS_P3 (1 << 3) -#define INTENS_C (1 << 31) -#define INTENS_MASK 0x8000000f /* Mask for writable bits */ - -/* - * EVTSEL: Event selection reg - */ -#define EVTSEL_MASK 0x7f /* Mask for writable bits */ - -/* - * SELECT: Counter selection reg - */ -#define SELECT_MASK 0x1f /* Mask for writable bits */ - -/* - * FLAG: counters overflow flag status reg - */ -#define FLAG_P0 (1 << 0) -#define FLAG_P1 (1 << 1) -#define FLAG_P2 (1 << 2) -#define FLAG_P3 (1 << 3) -#define FLAG_C (1 << 31) -#define FLAG_MASK 0x8000000f /* Mask for writable bits */ - - -int armv7_setup_pmu(void); -int armv7_start_pmu(void); -int armv7_stop_pmu(void); -int armv7_request_interrupts(const int *, int); -void armv7_release_interrupts(const int *, int); - -#endif diff --git a/arch/arm/oprofile/op_model_xscale.c b/arch/arm/oprofile/op_model_xscale.c deleted file mode 100644 index 1d34a02..0000000 --- a/arch/arm/oprofile/op_model_xscale.c +++ /dev/null @@ -1,444 +0,0 @@ -/** - * @file op_model_xscale.c - * XScale Performance Monitor Driver - * - * @remark Copyright 2000-2004 Deepak Saxena <dsaxena@mvista.com> - * @remark Copyright 2000-2004 MontaVista Software Inc - * @remark Copyright 2004 Dave Jiang <dave.jiang@intel.com> - * @remark Copyright 2004 Intel Corporation - * @remark Copyright 2004 Zwane Mwaikambo <zwane@arm.linux.org.uk> - * @remark Copyright 2004 OProfile Authors - * - * @remark Read the file COPYING - * - * @author Zwane Mwaikambo - */ - -/* #define DEBUG */ -#include <linux/types.h> -#include <linux/errno.h> -#include <linux/err.h> -#include <linux/sched.h> -#include <linux/oprofile.h> -#include <linux/interrupt.h> -#include <linux/irq.h> - -#include <asm/cputype.h> -#include <asm/pmu.h> - -#include "op_counter.h" -#include "op_arm_model.h" - -#define PMU_ENABLE 0x001 /* Enable counters */ -#define PMN_RESET 0x002 /* Reset event counters */ -#define CCNT_RESET 0x004 /* Reset clock counter */ -#define PMU_RESET (CCNT_RESET | PMN_RESET) -#define PMU_CNT64 0x008 /* Make CCNT count every 64th cycle */ - -/* - * Different types of events that can be counted by the XScale PMU - * as used by Oprofile userspace. Here primarily for documentation - * purposes. - */ - -#define EVT_ICACHE_MISS 0x00 -#define EVT_ICACHE_NO_DELIVER 0x01 -#define EVT_DATA_STALL 0x02 -#define EVT_ITLB_MISS 0x03 -#define EVT_DTLB_MISS 0x04 -#define EVT_BRANCH 0x05 -#define EVT_BRANCH_MISS 0x06 -#define EVT_INSTRUCTION 0x07 -#define EVT_DCACHE_FULL_STALL 0x08 -#define EVT_DCACHE_FULL_STALL_CONTIG 0x09 -#define EVT_DCACHE_ACCESS 0x0A -#define EVT_DCACHE_MISS 0x0B -#define EVT_DCACE_WRITE_BACK 0x0C -#define EVT_PC_CHANGED 0x0D -#define EVT_BCU_REQUEST 0x10 -#define EVT_BCU_FULL 0x11 -#define EVT_BCU_DRAIN 0x12 -#define EVT_BCU_ECC_NO_ELOG 0x14 -#define EVT_BCU_1_BIT_ERR 0x15 -#define EVT_RMW 0x16 -/* EVT_CCNT is not hardware defined */ -#define EVT_CCNT 0xFE -#define EVT_UNUSED 0xFF - -struct pmu_counter { - volatile unsigned long ovf; - unsigned long reset_counter; -}; - -enum { CCNT, PMN0, PMN1, PMN2, PMN3, MAX_COUNTERS }; - -static struct pmu_counter results[MAX_COUNTERS]; - -/* - * There are two versions of the PMU in current XScale processors - * with differing register layouts and number of performance counters. - * e.g. IOP32x is xsc1 whilst IOP33x is xsc2. - * We detect which register layout to use in xscale_detect_pmu() - */ -enum { PMU_XSC1, PMU_XSC2 }; - -struct pmu_type { - int id; - char *name; - int num_counters; - unsigned int int_enable; - unsigned int cnt_ovf[MAX_COUNTERS]; - unsigned int int_mask[MAX_COUNTERS]; -}; - -static struct pmu_type pmu_parms[] = { - { - .id = PMU_XSC1, - .name = "arm/xscale1", - .num_counters = 3, - .int_mask = { [PMN0] = 0x10, [PMN1] = 0x20, - [CCNT] = 0x40 }, - .cnt_ovf = { [CCNT] = 0x400, [PMN0] = 0x100, - [PMN1] = 0x200}, - }, - { - .id = PMU_XSC2, - .name = "arm/xscale2", - .num_counters = 5, - .int_mask = { [CCNT] = 0x01, [PMN0] = 0x02, - [PMN1] = 0x04, [PMN2] = 0x08, - [PMN3] = 0x10 }, - .cnt_ovf = { [CCNT] = 0x01, [PMN0] = 0x02, - [PMN1] = 0x04, [PMN2] = 0x08, - [PMN3] = 0x10 }, - }, -}; - -static struct pmu_type *pmu; - -static void write_pmnc(u32 val) -{ - if (pmu->id == PMU_XSC1) { - /* upper 4bits and 7, 11 are write-as-0 */ - val &= 0xffff77f; - __asm__ __volatile__ ("mcr p14, 0, %0, c0, c0, 0" : : "r" (val)); - } else { - /* bits 4-23 are write-as-0, 24-31 are write ignored */ - val &= 0xf; - __asm__ __volatile__ ("mcr p14, 0, %0, c0, c1, 0" : : "r" (val)); - } -} - -static u32 read_pmnc(void) -{ - u32 val; - - if (pmu->id == PMU_XSC1) - __asm__ __volatile__ ("mrc p14, 0, %0, c0, c0, 0" : "=r" (val)); - else { - __asm__ __volatile__ ("mrc p14, 0, %0, c0, c1, 0" : "=r" (val)); - /* bits 1-2 and 4-23 are read-unpredictable */ - val &= 0xff000009; - } - - return val; -} - -static u32 __xsc1_read_counter(int counter) -{ - u32 val = 0; - - switch (counter) { - case CCNT: - __asm__ __volatile__ ("mrc p14, 0, %0, c1, c0, 0" : "=r" (val)); - break; - case PMN0: - __asm__ __volatile__ ("mrc p14, 0, %0, c2, c0, 0" : "=r" (val)); - break; - case PMN1: - __asm__ __volatile__ ("mrc p14, 0, %0, c3, c0, 0" : "=r" (val)); - break; - } - return val; -} - -static u32 __xsc2_read_counter(int counter) -{ - u32 val = 0; - - switch (counter) { - case CCNT: - __asm__ __volatile__ ("mrc p14, 0, %0, c1, c1, 0" : "=r" (val)); - break; - case PMN0: - __asm__ __volatile__ ("mrc p14, 0, %0, c0, c2, 0" : "=r" (val)); - break; - case PMN1: - __asm__ __volatile__ ("mrc p14, 0, %0, c1, c2, 0" : "=r" (val)); - break; - case PMN2: - __asm__ __volatile__ ("mrc p14, 0, %0, c2, c2, 0" : "=r" (val)); - break; - case PMN3: - __asm__ __volatile__ ("mrc p14, 0, %0, c3, c2, 0" : "=r" (val)); - break; - } - return val; -} - -static u32 read_counter(int counter) -{ - u32 val; - - if (pmu->id == PMU_XSC1) - val = __xsc1_read_counter(counter); - else - val = __xsc2_read_counter(counter); - - return val; -} - -static void __xsc1_write_counter(int counter, u32 val) -{ - switch (counter) { - case CCNT: - __asm__ __volatile__ ("mcr p14, 0, %0, c1, c0, 0" : : "r" (val)); - break; - case PMN0: - __asm__ __volatile__ ("mcr p14, 0, %0, c2, c0, 0" : : "r" (val)); - break; - case PMN1: - __asm__ __volatile__ ("mcr p14, 0, %0, c3, c0, 0" : : "r" (val)); - break; - } -} - -static void __xsc2_write_counter(int counter, u32 val) -{ - switch (counter) { - case CCNT: - __asm__ __volatile__ ("mcr p14, 0, %0, c1, c1, 0" : : "r" (val)); - break; - case PMN0: - __asm__ __volatile__ ("mcr p14, 0, %0, c0, c2, 0" : : "r" (val)); - break; - case PMN1: - __asm__ __volatile__ ("mcr p14, 0, %0, c1, c2, 0" : : "r" (val)); - break; - case PMN2: - __asm__ __volatile__ ("mcr p14, 0, %0, c2, c2, 0" : : "r" (val)); - break; - case PMN3: - __asm__ __volatile__ ("mcr p14, 0, %0, c3, c2, 0" : : "r" (val)); - break; - } -} - -static void write_counter(int counter, u32 val) -{ - if (pmu->id == PMU_XSC1) - __xsc1_write_counter(counter, val); - else - __xsc2_write_counter(counter, val); -} - -static int xscale_setup_ctrs(void) -{ - u32 evtsel, pmnc; - int i; - - for (i = CCNT; i < MAX_COUNTERS; i++) { - if (counter_config[i].enabled) - continue; - - counter_config[i].event = EVT_UNUSED; - } - - switch (pmu->id) { - case PMU_XSC1: - pmnc = (counter_config[PMN1].event << 20) | (counter_config[PMN0].event << 12); - pr_debug("xscale_setup_ctrs: pmnc: %#08x\n", pmnc); - write_pmnc(pmnc); - break; - - case PMU_XSC2: - evtsel = counter_config[PMN0].event | (counter_config[PMN1].event << 8) | - (counter_config[PMN2].event << 16) | (counter_config[PMN3].event << 24); - - pr_debug("xscale_setup_ctrs: evtsel %#08x\n", evtsel); - __asm__ __volatile__ ("mcr p14, 0, %0, c8, c1, 0" : : "r" (evtsel)); - break; - } - - for (i = CCNT; i < MAX_COUNTERS; i++) { - if (counter_config[i].event == EVT_UNUSED) { - counter_config[i].event = 0; - pmu->int_enable &= ~pmu->int_mask[i]; - continue; - } - - results[i].reset_counter = counter_config[i].count; - write_counter(i, -(u32)counter_config[i].count); - pmu->int_enable |= pmu->int_mask[i]; - pr_debug("xscale_setup_ctrs: counter%d %#08x from %#08lx\n", i, - read_counter(i), counter_config[i].count); - } - - return 0; -} - -static void inline __xsc1_check_ctrs(void) -{ - int i; - u32 pmnc = read_pmnc(); - - /* NOTE: there's an A stepping errata that states if an overflow */ - /* bit already exists and another occurs, the previous */ - /* Overflow bit gets cleared. There's no workaround. */ - /* Fixed in B stepping or later */ - - /* Write the value back to clear the overflow flags. Overflow */ - /* flags remain in pmnc for use below */ - write_pmnc(pmnc & ~PMU_ENABLE); - - for (i = CCNT; i <= PMN1; i++) { - if (!(pmu->int_mask[i] & pmu->int_enable)) - continue; - - if (pmnc & pmu->cnt_ovf[i]) - results[i].ovf++; - } -} - -static void inline __xsc2_check_ctrs(void) -{ - int i; - u32 flag = 0, pmnc = read_pmnc(); - - pmnc &= ~PMU_ENABLE; - write_pmnc(pmnc); - - /* read overflow flag register */ - __asm__ __volatile__ ("mrc p14, 0, %0, c5, c1, 0" : "=r" (flag)); - - for (i = CCNT; i <= PMN3; i++) { - if (!(pmu->int_mask[i] & pmu->int_enable)) - continue; - - if (flag & pmu->cnt_ovf[i]) - results[i].ovf++; - } - - /* writeback clears overflow bits */ - __asm__ __volatile__ ("mcr p14, 0, %0, c5, c1, 0" : : "r" (flag)); -} - -static irqreturn_t xscale_pmu_interrupt(int irq, void *arg) -{ - int i; - u32 pmnc; - - if (pmu->id == PMU_XSC1) - __xsc1_check_ctrs(); - else - __xsc2_check_ctrs(); - - for (i = CCNT; i < MAX_COUNTERS; i++) { - if (!results[i].ovf) - continue; - - write_counter(i, -(u32)results[i].reset_counter); - oprofile_add_sample(get_irq_regs(), i); - results[i].ovf--; - } - - pmnc = read_pmnc() | PMU_ENABLE; - write_pmnc(pmnc); - - return IRQ_HANDLED; -} - -static const struct pmu_irqs *pmu_irqs; - -static void xscale_pmu_stop(void) -{ - u32 pmnc = read_pmnc(); - - pmnc &= ~PMU_ENABLE; - write_pmnc(pmnc); - - free_irq(pmu_irqs->irqs[0], results); - release_pmu(pmu_irqs); - pmu_irqs = NULL; -} - -static int xscale_pmu_start(void) -{ - int ret; - u32 pmnc; - - pmu_irqs = reserve_pmu(); - if (IS_ERR(pmu_irqs)) - return PTR_ERR(pmu_irqs); - - pmnc = read_pmnc(); - - ret = request_irq(pmu_irqs->irqs[0], xscale_pmu_interrupt, - IRQF_DISABLED, "XScale PMU", (void *)results); - - if (ret < 0) { - printk(KERN_ERR "oprofile: unable to request IRQ%d for XScale PMU\n", - pmu_irqs->irqs[0]); - release_pmu(pmu_irqs); - pmu_irqs = NULL; - return ret; - } - - if (pmu->id == PMU_XSC1) - pmnc |= pmu->int_enable; - else { - __asm__ __volatile__ ("mcr p14, 0, %0, c4, c1, 0" : : "r" (pmu->int_enable)); - pmnc &= ~PMU_CNT64; - } - - pmnc |= PMU_ENABLE; - write_pmnc(pmnc); - pr_debug("xscale_pmu_start: pmnc: %#08x mask: %08x\n", pmnc, pmu->int_enable); - return 0; -} - -static int xscale_detect_pmu(void) -{ - int ret = 0; - u32 id; - - id = (read_cpuid(CPUID_ID) >> 13) & 0x7; - - switch (id) { - case 1: - pmu = &pmu_parms[PMU_XSC1]; - break; - case 2: - pmu = &pmu_parms[PMU_XSC2]; - break; - default: - ret = -ENODEV; - break; - } - - if (!ret) { - op_xscale_spec.name = pmu->name; - op_xscale_spec.num_counters = pmu->num_counters; - pr_debug("xscale_detect_pmu: detected %s PMU\n", pmu->name); - } - - return ret; -} - -struct op_arm_model_spec op_xscale_spec = { - .init = xscale_detect_pmu, - .setup_ctrs = xscale_setup_ctrs, - .start = xscale_pmu_start, - .stop = xscale_pmu_stop, -}; - -- 1.6.3.3 ^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules 2010-03-10 10:41 ` [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules Will Deacon 2010-03-10 10:41 ` [PATCH 5/6] ARM: oprofile: use perf-events framework as backend Will Deacon @ 2010-03-10 10:56 ` Peter Zijlstra 2010-03-10 11:47 ` Will Deacon 1 sibling, 1 reply; 19+ messages in thread From: Peter Zijlstra @ 2010-03-10 10:56 UTC (permalink / raw) To: linux-arm-kernel On Wed, 2010-03-10 at 10:41 +0000, Will Deacon wrote: > The perf_event_enable and perf_event_disable functions are used to > control the activation of perf-events for a given context or CPU. > > This patch exports these symbols so that they can be used by kernel > modules. Without these symbols, an event must be destroyed and recreated > to disable or enable it respectively. The maximum number of perf-events > is also made available to modules via the perf_get_max_events function. Where is the module that uses all this? ^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules 2010-03-10 10:56 ` [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules Peter Zijlstra @ 2010-03-10 11:47 ` Will Deacon 2010-03-10 14:30 ` Peter Zijlstra 0 siblings, 1 reply; 19+ messages in thread From: Will Deacon @ 2010-03-10 11:47 UTC (permalink / raw) To: linux-arm-kernel Hello Peter, > On Wed, 2010-03-10 at 10:41 +0000, Will Deacon wrote: > > The perf_event_enable and perf_event_disable functions are used to > > control the activation of perf-events for a given context or CPU. > > > > This patch exports these symbols so that they can be used by kernel > > modules. Without these symbols, an event must be destroyed and recreated > > to disable or enable it respectively. The maximum number of perf-events > > is also made available to modules via the perf_get_max_events function. > > Where is the module that uses all this? It's part of the same patchset: http://lists.infradead.org/pipermail/linux-arm-kernel/2010-March/011126.html The idea is to use perf-events as the backend for OProfile to avoid replicating the PMU accessor code. Cheers, Will ^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules 2010-03-10 11:47 ` Will Deacon @ 2010-03-10 14:30 ` Peter Zijlstra 2010-03-10 14:41 ` Will Deacon 2010-03-10 15:01 ` Robert Richter 0 siblings, 2 replies; 19+ messages in thread From: Peter Zijlstra @ 2010-03-10 14:30 UTC (permalink / raw) To: linux-arm-kernel On Wed, 2010-03-10 at 11:47 +0000, Will Deacon wrote: > Hello Peter, > > > On Wed, 2010-03-10 at 10:41 +0000, Will Deacon wrote: > > > The perf_event_enable and perf_event_disable functions are used to > > > control the activation of perf-events for a given context or CPU. > > > > > > This patch exports these symbols so that they can be used by kernel > > > modules. Without these symbols, an event must be destroyed and recreated > > > to disable or enable it respectively. The maximum number of perf-events > > > is also made available to modules via the perf_get_max_events function. > > > > Where is the module that uses all this? > > It's part of the same patchset: > > http://lists.infradead.org/pipermail/linux-arm-kernel/2010-March/011126.html > > The idea is to use perf-events as the backend for OProfile to avoid replicating > the PMU accessor code. Ah, very nice. Can something like that be done in general, like an arch-independent oprofile backend using perf, or is there too much arch-specific bits to the oprofile stuff? I know Robert talked about doing something like this, I just have no clue how arch specific this would be. ^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules 2010-03-10 14:30 ` Peter Zijlstra @ 2010-03-10 14:41 ` Will Deacon 2010-03-10 15:01 ` Robert Richter 1 sibling, 0 replies; 19+ messages in thread From: Will Deacon @ 2010-03-10 14:41 UTC (permalink / raw) To: linux-arm-kernel Hi Peter, > Ah, very nice. Can something like that be done in general, like an > arch-independent oprofile backend using perf, or is there too much > arch-specific bits to the oprofile stuff? In theory, it should be possible but it would require that: (a) all the architectures that have OProfile support have equivalent support in Perf (b) the OProfile userspace tools would need tweaking so that the kernel interface: i.) is identical across different architectures ii.) has a 1:1 mapping with perf. That is, OProfile tools must agree with perf on things like number of counters and PMU name [which I don't think the generic perf code needs to care about]. > I know Robert talked about doing something like this, I just have no > clue how arch specific this would be. I think it would be a nightmare to fix up all the edge-cases that each architecture has at the moment. If, in the long run, each architecture ends up defining their own oprofile <-> perf interface then it would make sense to look at merging them at the higher level. Will ^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules 2010-03-10 14:30 ` Peter Zijlstra 2010-03-10 14:41 ` Will Deacon @ 2010-03-10 15:01 ` Robert Richter 1 sibling, 0 replies; 19+ messages in thread From: Robert Richter @ 2010-03-10 15:01 UTC (permalink / raw) To: linux-arm-kernel On 10.03.10 15:30:42, Peter Zijlstra wrote: > On Wed, 2010-03-10 at 11:47 +0000, Will Deacon wrote: > > Hello Peter, > > > > > On Wed, 2010-03-10 at 10:41 +0000, Will Deacon wrote: > > > > The perf_event_enable and perf_event_disable functions are used to > > > > control the activation of perf-events for a given context or CPU. > > > > > > > > This patch exports these symbols so that they can be used by kernel > > > > modules. Without these symbols, an event must be destroyed and recreated > > > > to disable or enable it respectively. The maximum number of perf-events > > > > is also made available to modules via the perf_get_max_events function. > > > > > > Where is the module that uses all this? > > > > It's part of the same patchset: > > > > http://lists.infradead.org/pipermail/linux-arm-kernel/2010-March/011126.html > > > > The idea is to use perf-events as the backend for OProfile to avoid replicating > > the PMU accessor code. > > Ah, very nice. Can something like that be done in general, like an > arch-independent oprofile backend using perf, or is there too much > arch-specific bits to the oprofile stuff? > > I know Robert talked about doing something like this, I just have no > clue how arch specific this would be. Thanks for cc'ing me. I like this approach. Will take a look at the patches. Yes, for implementing oprofile on top of perf we have to touch every architecture separately to be backward compatible. -Robert -- Advanced Micro Devices, Inc. Operating System Research Center email: robert.richter at amd.com ^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU 2010-03-10 10:41 ` [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU Will Deacon 2010-03-10 10:41 ` [PATCH 3/6] ARM: perf-events: add support for xscale PMUs Will Deacon @ 2010-03-10 11:03 ` Jamie Iles 1 sibling, 0 replies; 19+ messages in thread From: Jamie Iles @ 2010-03-10 11:03 UTC (permalink / raw) To: linux-arm-kernel On Wed, Mar 10, 2010 at 10:41:26AM +0000, Will Deacon wrote: > The ARM perf-events framework provides support for a number of different > PMUs using struct arm_pmu. The char *name field of this struct can be > used to identify the PMU, but this is cumbersome if used outside of perf. > > This patch replaces the name string for a PMU with an enum, which holds > a unique ID for the PMU being represented. This ID can be used to index > an array of names within perf, so no functionality is lost. The presence > of the ID field, allows other kernel subsystems [currently oprofile] to > use their own mappings for the PMU name. > > Cc: Jamie Iles <jamie.iles@picochip.com> > Cc: Jean Pihet <jpihet@mvista.com> > Signed-off-by: Will Deacon <will.deacon@arm.com> Hi Will, Looks good to me. Acked-by: Jamie Iles <jamie.iles@picochip.com> Jamie ^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c 2010-03-10 10:41 ` [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c Will Deacon 2010-03-10 10:41 ` [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU Will Deacon @ 2010-03-10 21:59 ` Tony Lindgren 2010-03-10 22:27 ` Albin Tonnerre 1 sibling, 1 reply; 19+ messages in thread From: Tony Lindgren @ 2010-03-10 21:59 UTC (permalink / raw) To: linux-arm-kernel * Will Deacon <will.deacon@arm.com> [100310 02:46]: > Add the PMU IRQs for the Realview boards to kernel/pmu.c so that > they can be used by the perf-events framework. > > Signed-off-by: Will Deacon <will.deacon@arm.com> > --- > arch/arm/kernel/pmu.c | 14 ++++++++++++++ > 1 files changed, 14 insertions(+), 0 deletions(-) > > diff --git a/arch/arm/kernel/pmu.c b/arch/arm/kernel/pmu.c > index a124312..11a321e 100644 > --- a/arch/arm/kernel/pmu.c > +++ b/arch/arm/kernel/pmu.c > @@ -36,6 +36,20 @@ static const int irqs[] = { > IRQ_EB11MP_PMU_CPU1, > IRQ_EB11MP_PMU_CPU2, > IRQ_EB11MP_PMU_CPU3, > +#elif defined(CONFIG_MACH_REALVIEW_PB1176) > + IRQ_DC1176_CORE_PMU, > +#elif defined(CONFIG_MACH_REALVIEW_PB11MP) > + IRQ_TC11MP_PMU_CPU0, > + IRQ_TC11MP_PMU_CPU1, > + IRQ_TC11MP_PMU_CPU2, > + IRQ_TC11MP_PMU_CPU3, > +#elif defined(CONFIG_MACH_REALVIEW_PBA8) > + IRQ_PBA8_PMU, > +#elif defined(CONFIG_MACH_REALVIEW_PBX) > + IRQ_PBX_PMU_CPU0, > + IRQ_PBX_PMU_CPU1, > + IRQ_PBX_PMU_CPU2, > + IRQ_PBX_PMU_CPU3, > #elif defined(CONFIG_ARCH_OMAP3) > INT_34XX_BENCH_MPU_EMUL, > #elif defined(CONFIG_ARCH_IOP32X) Just a generic comment.. It would be nice to get rid of the ifdef eliffery here for the interrupt. At least we can now compile omap2 + 3 + 4 into the same binary, so stuff like this caused mysterious errors easily. Regards, Tony ^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c 2010-03-10 21:59 ` [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c Tony Lindgren @ 2010-03-10 22:27 ` Albin Tonnerre 2010-03-10 22:34 ` Tony Lindgren 0 siblings, 1 reply; 19+ messages in thread From: Albin Tonnerre @ 2010-03-10 22:27 UTC (permalink / raw) To: linux-arm-kernel On Wed, 10 Mar 2010 13:59 -0800, Tony Lindgren wrote : > * Will Deacon <will.deacon@arm.com> [100310 02:46]: > > Add the PMU IRQs for the Realview boards to kernel/pmu.c so that > > they can be used by the perf-events framework. > > > > Signed-off-by: Will Deacon <will.deacon@arm.com> > > --- > > arch/arm/kernel/pmu.c | 14 ++++++++++++++ > > 1 files changed, 14 insertions(+), 0 deletions(-) > > > > diff --git a/arch/arm/kernel/pmu.c b/arch/arm/kernel/pmu.c > > index a124312..11a321e 100644 > > --- a/arch/arm/kernel/pmu.c > > +++ b/arch/arm/kernel/pmu.c > > @@ -36,6 +36,20 @@ static const int irqs[] = { > > IRQ_EB11MP_PMU_CPU1, > > IRQ_EB11MP_PMU_CPU2, > > IRQ_EB11MP_PMU_CPU3, > > +#elif defined(CONFIG_MACH_REALVIEW_PB1176) > > + IRQ_DC1176_CORE_PMU, > > +#elif defined(CONFIG_MACH_REALVIEW_PB11MP) > > + IRQ_TC11MP_PMU_CPU0, > > + IRQ_TC11MP_PMU_CPU1, > > + IRQ_TC11MP_PMU_CPU2, > > + IRQ_TC11MP_PMU_CPU3, > > +#elif defined(CONFIG_MACH_REALVIEW_PBA8) > > + IRQ_PBA8_PMU, > > +#elif defined(CONFIG_MACH_REALVIEW_PBX) > > + IRQ_PBX_PMU_CPU0, > > + IRQ_PBX_PMU_CPU1, > > + IRQ_PBX_PMU_CPU2, > > + IRQ_PBX_PMU_CPU3, > > #elif defined(CONFIG_ARCH_OMAP3) > > INT_34XX_BENCH_MPU_EMUL, > > #elif defined(CONFIG_ARCH_IOP32X) > > Just a generic comment.. It would be nice to get rid of the > ifdef eliffery here for the interrupt. At least we can now > compile omap2 + 3 + 4 into the same binary, so stuff like > this caused mysterious errors easily. That's the purpose of another patch series Will has submitted as an RFC, see '[RFC PATCH 0/2] ARM: pmu: provide a registration mechanism for IRQs' Cheers, -- Albin Tonnerre ^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c 2010-03-10 22:27 ` Albin Tonnerre @ 2010-03-10 22:34 ` Tony Lindgren 0 siblings, 0 replies; 19+ messages in thread From: Tony Lindgren @ 2010-03-10 22:34 UTC (permalink / raw) To: linux-arm-kernel * Albin Tonnerre <albin.tonnerre@gmail.com> [100310 14:22]: > On Wed, 10 Mar 2010 13:59 -0800, Tony Lindgren wrote : > > * Will Deacon <will.deacon@arm.com> [100310 02:46]: > > > Add the PMU IRQs for the Realview boards to kernel/pmu.c so that > > > they can be used by the perf-events framework. > > > > > > Signed-off-by: Will Deacon <will.deacon@arm.com> > > > --- > > > arch/arm/kernel/pmu.c | 14 ++++++++++++++ > > > 1 files changed, 14 insertions(+), 0 deletions(-) > > > > > > diff --git a/arch/arm/kernel/pmu.c b/arch/arm/kernel/pmu.c > > > index a124312..11a321e 100644 > > > --- a/arch/arm/kernel/pmu.c > > > +++ b/arch/arm/kernel/pmu.c > > > @@ -36,6 +36,20 @@ static const int irqs[] = { > > > IRQ_EB11MP_PMU_CPU1, > > > IRQ_EB11MP_PMU_CPU2, > > > IRQ_EB11MP_PMU_CPU3, > > > +#elif defined(CONFIG_MACH_REALVIEW_PB1176) > > > + IRQ_DC1176_CORE_PMU, > > > +#elif defined(CONFIG_MACH_REALVIEW_PB11MP) > > > + IRQ_TC11MP_PMU_CPU0, > > > + IRQ_TC11MP_PMU_CPU1, > > > + IRQ_TC11MP_PMU_CPU2, > > > + IRQ_TC11MP_PMU_CPU3, > > > +#elif defined(CONFIG_MACH_REALVIEW_PBA8) > > > + IRQ_PBA8_PMU, > > > +#elif defined(CONFIG_MACH_REALVIEW_PBX) > > > + IRQ_PBX_PMU_CPU0, > > > + IRQ_PBX_PMU_CPU1, > > > + IRQ_PBX_PMU_CPU2, > > > + IRQ_PBX_PMU_CPU3, > > > #elif defined(CONFIG_ARCH_OMAP3) > > > INT_34XX_BENCH_MPU_EMUL, > > > #elif defined(CONFIG_ARCH_IOP32X) > > > > Just a generic comment.. It would be nice to get rid of the > > ifdef eliffery here for the interrupt. At least we can now > > compile omap2 + 3 + 4 into the same binary, so stuff like > > this caused mysterious errors easily. > > That's the purpose of another patch series Will has submitted as an RFC, > see '[RFC PATCH 0/2] ARM: pmu: provide a registration mechanism for IRQs' Oh, great! Sorry I did not notice that & thanks for doing that. Regards, Tony ^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 0/6] ARM: oprofile: use perf-events framework as backend @ 2010-02-25 18:56 Will Deacon 2010-02-25 18:56 ` [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c Will Deacon 0 siblings, 1 reply; 19+ messages in thread From: Will Deacon @ 2010-02-25 18:56 UTC (permalink / raw) To: linux-arm-kernel There are currently two hardware performance monitoring subsystems in the kernel for ARM: OProfile and perf-events. This creates the following problems: 1.) Duplicate PMU accessor code. Inevitable code drift may lead to bugs in one framework that are fixed in the other. 2.) Locking issues. OProfile doesn't reprogram hardware counters between profiling runs if the events to be monitored have not been changed. This means that other profiling frameworks cannot use the counters if OProfile is in use. 3.) Due to differences in the two frameworks, it may not be possible to compare the results obtained by OProfile with those obtained by perf. This patch series removes the OProfile PMU driver code and replaces it with calls to perf, therefore resolving the issues mentioned above. To preserve compatibility with existing OProfile tools, drivers for XScale v1 and v2 PMUs are added to the perf backend. The only userspace-visible change is the lack of SCU counter support for 11MPCore. This is currently unsupported by OProfile userspace tools anyway and therefore shouldn't cause any problems. Tested on dual-core A9MP [Realview PBX], quad-core 11MPCore [Realview PB11MP] and XScale PXA270 [Gumstix Verdex XL6P]. Patch taken against 2.6.33 + perf-events patches in Russell's branch: ARM: 5890/1: Fix incorrect Realview board IRQs for L220 and PMU ARM: 5899/2: arm: provide a mechanism to reserve performance counters ARM: 5901/2: arm/oprofile: reserve the PMU when starting ARM: 5900/2: arm: enable support for software perf events ARM: 5902/4: arm/perfevents: implement perf event support for ARMv6 ARM: 5903/1: arm/perfevents: add support for ARMv7 Will Deacon (6): ARM: perf-events: add Realview PMU IRQs to pmu.c ARM: perf-events: use numeric ID to identify PMU ARM: perf-events: add support for xscale PMUs perf-events: export enable/disable event symbols to kernel modules ARM: oprofile: use perf-events framework as backend ARM: oprofile: remove old files and update KConfig arch/arm/Kconfig | 26 +- arch/arm/include/asm/perf_event.h | 12 + arch/arm/kernel/perf_event.c | 862 ++++++++++++++++++++++++++++++- arch/arm/kernel/pmu.c | 14 + arch/arm/oprofile/Makefile | 7 +- arch/arm/oprofile/backtrace.c | 83 --- arch/arm/oprofile/common.c | 356 +++++++++++-- arch/arm/oprofile/op_arm_model.h | 35 -- arch/arm/oprofile/op_counter.h | 27 - arch/arm/oprofile/op_model_arm11_core.c | 162 ------ arch/arm/oprofile/op_model_arm11_core.h | 45 -- arch/arm/oprofile/op_model_mpcore.c | 306 ----------- arch/arm/oprofile/op_model_mpcore.h | 61 --- arch/arm/oprofile/op_model_v6.c | 78 --- arch/arm/oprofile/op_model_v7.c | 415 --------------- arch/arm/oprofile/op_model_v7.h | 103 ---- arch/arm/oprofile/op_model_xscale.c | 444 ---------------- include/linux/perf_event.h | 1 + kernel/perf_event.c | 8 + 19 files changed, 1192 insertions(+), 1853 deletions(-) delete mode 100644 arch/arm/oprofile/backtrace.c delete mode 100644 arch/arm/oprofile/op_arm_model.h delete mode 100644 arch/arm/oprofile/op_counter.h delete mode 100644 arch/arm/oprofile/op_model_arm11_core.c delete mode 100644 arch/arm/oprofile/op_model_arm11_core.h delete mode 100644 arch/arm/oprofile/op_model_mpcore.c delete mode 100644 arch/arm/oprofile/op_model_mpcore.h delete mode 100644 arch/arm/oprofile/op_model_v6.c delete mode 100644 arch/arm/oprofile/op_model_v7.c delete mode 100644 arch/arm/oprofile/op_model_v7.h delete mode 100644 arch/arm/oprofile/op_model_xscale.c ^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c 2010-02-25 18:56 [PATCH 0/6] ARM: oprofile: use perf-events framework as backend Will Deacon @ 2010-02-25 18:56 ` Will Deacon 2010-02-25 18:56 ` [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU Will Deacon 0 siblings, 1 reply; 19+ messages in thread From: Will Deacon @ 2010-02-25 18:56 UTC (permalink / raw) To: linux-arm-kernel Add the PMU IRQs for the Realview boards to kernel/pmu.c so that they can be used by the perf-events framework. Signed-off-by: Will Deacon <will.deacon@arm.com> --- arch/arm/kernel/pmu.c | 14 ++++++++++++++ 1 files changed, 14 insertions(+), 0 deletions(-) diff --git a/arch/arm/kernel/pmu.c b/arch/arm/kernel/pmu.c index a124312..11a321e 100644 --- a/arch/arm/kernel/pmu.c +++ b/arch/arm/kernel/pmu.c @@ -36,6 +36,20 @@ static const int irqs[] = { IRQ_EB11MP_PMU_CPU1, IRQ_EB11MP_PMU_CPU2, IRQ_EB11MP_PMU_CPU3, +#elif defined(CONFIG_MACH_REALVIEW_PB1176) + IRQ_DC1176_CORE_PMU, +#elif defined(CONFIG_MACH_REALVIEW_PB11MP) + IRQ_TC11MP_PMU_CPU0, + IRQ_TC11MP_PMU_CPU1, + IRQ_TC11MP_PMU_CPU2, + IRQ_TC11MP_PMU_CPU3, +#elif defined(CONFIG_MACH_REALVIEW_PBA8) + IRQ_PBA8_PMU, +#elif defined(CONFIG_MACH_REALVIEW_PBX) + IRQ_PBX_PMU_CPU0, + IRQ_PBX_PMU_CPU1, + IRQ_PBX_PMU_CPU2, + IRQ_PBX_PMU_CPU3, #elif defined(CONFIG_ARCH_OMAP3) INT_34XX_BENCH_MPU_EMUL, #elif defined(CONFIG_ARCH_IOP32X) -- 1.6.3.3 ^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU 2010-02-25 18:56 ` [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c Will Deacon @ 2010-02-25 18:56 ` Will Deacon 2010-02-25 18:56 ` [PATCH 3/6] ARM: perf-events: add support for xscale PMUs Will Deacon 0 siblings, 1 reply; 19+ messages in thread From: Will Deacon @ 2010-02-25 18:56 UTC (permalink / raw) To: linux-arm-kernel The ARM perf-events framework provides support for a number of different PMUs using struct arm_pmu. The char *name field of this struct can be used to identify the PMU, but this is cumbersome if used outside of perf. This patch replaces the name string for a PMU with an int, which holds a unique ID for the PMU being represented. This ID can be used to index an array of names within perf, so no functionality is lost. The presence of the ID field, allows other kernel subsystems [currently oprofile] to use their own mappings for the PMU name. Cc: Jamie Iles <jamie.iles@picochip.com> Signed-off-by: Will Deacon <will.deacon@arm.com> --- arch/arm/include/asm/perf_event.h | 12 +++++++++++ arch/arm/kernel/perf_event.c | 39 +++++++++++++++++++++++++++--------- 2 files changed, 41 insertions(+), 10 deletions(-) diff --git a/arch/arm/include/asm/perf_event.h b/arch/arm/include/asm/perf_event.h index 49e3049..925b9a4 100644 --- a/arch/arm/include/asm/perf_event.h +++ b/arch/arm/include/asm/perf_event.h @@ -28,4 +28,16 @@ set_perf_event_pending(void) * same indexes here for consistency. */ #define PERF_EVENT_INDEX_OFFSET 1 +/* ARM perf PMU IDs for use by internal perf clients. */ +#define ARM_PERF_PMU_ID_XSCALE1 0 +#define ARM_PERF_PMU_ID_XSCALE2 (ARM_PERF_PMU_ID_XSCALE1 + 1) +#define ARM_PERF_PMU_ID_V6 (ARM_PERF_PMU_ID_XSCALE2 + 1) +#define ARM_PERF_PMU_ID_V6MP (ARM_PERF_PMU_ID_V6 + 1) +#define ARM_PERF_PMU_ID_CA8 (ARM_PERF_PMU_ID_V6MP + 1) +#define ARM_PERF_PMU_ID_CA9 (ARM_PERF_PMU_ID_CA8 + 1) +#define ARM_NUM_PMU_IDS (ARM_PERF_PMU_ID_CA9 + 1) + +extern int +armpmu_get_pmu_id(void); + #endif /* __ARM_PERF_EVENT_H__ */ diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c index c54ceb3..51d9711 100644 --- a/arch/arm/kernel/perf_event.c +++ b/arch/arm/kernel/perf_event.c @@ -16,6 +16,7 @@ #include <linux/interrupt.h> #include <linux/kernel.h> +#include <linux/module.h> #include <linux/perf_event.h> #include <linux/spinlock.h> #include <linux/uaccess.h> @@ -67,8 +68,18 @@ struct cpu_hw_events { }; DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events); +/* PMU names. */ +static const char *arm_pmu_names[ARM_NUM_PMU_IDS] = { + "xscale1", + "xscale2", + "v6", + "v6mpcore", + "ARMv7 Cortex-A8", + "ARMv7 Cortex-A9", +}; + struct arm_pmu { - char *name; + int id; irqreturn_t (*handle_irq)(int irq_num, void *dev); void (*enable)(struct hw_perf_event *evt, int idx); void (*disable)(struct hw_perf_event *evt, int idx); @@ -87,6 +98,18 @@ struct arm_pmu { /* Set at runtime when we know what CPU type we are. */ static const struct arm_pmu *armpmu; +int +armpmu_get_pmu_id(void) +{ + int id = -ENODEV; + + if (armpmu != NULL) + id = armpmu->id; + + return id; +} +EXPORT_SYMBOL_GPL(armpmu_get_pmu_id); + #define HW_OP_UNSUPPORTED 0xFFFF #define C(_x) \ @@ -1143,7 +1166,7 @@ armv6mpcore_pmu_disable_event(struct hw_perf_event *hwc, } static const struct arm_pmu armv6pmu = { - .name = "v6", + .id = ARM_PERF_PMU_ID_V6, .handle_irq = armv6pmu_handle_irq, .enable = armv6pmu_enable_event, .disable = armv6pmu_disable_event, @@ -1166,7 +1189,7 @@ static const struct arm_pmu armv6pmu = { * reset the period and enable the interrupt reporting. */ static const struct arm_pmu armv6mpcore_pmu = { - .name = "v6mpcore", + .id = ARM_PERF_PMU_ID_V6MP, .handle_irq = armv6pmu_handle_irq, .enable = armv6pmu_enable_event, .disable = armv6mpcore_pmu_disable_event, @@ -1196,10 +1219,6 @@ static const struct arm_pmu armv6mpcore_pmu = { * counter and all 4 performance counters together can be reset separately. */ -#define ARMV7_PMU_CORTEX_A8_NAME "ARMv7 Cortex-A8" - -#define ARMV7_PMU_CORTEX_A9_NAME "ARMv7 Cortex-A9" - /* Common ARMv7 event types */ enum armv7_perf_types { ARMV7_PERFCTR_PMNC_SW_INCR = 0x00, @@ -2104,7 +2123,7 @@ init_hw_perf_events(void) perf_max_events = armv6mpcore_pmu.num_events; break; case 0xC080: /* Cortex-A8 */ - armv7pmu.name = ARMV7_PMU_CORTEX_A8_NAME; + armv7pmu.id = ARM_PERF_PMU_ID_CA8; memcpy(armpmu_perf_cache_map, armv7_a8_perf_cache_map, sizeof(armv7_a8_perf_cache_map)); armv7pmu.event_map = armv7_a8_pmu_event_map; @@ -2116,7 +2135,7 @@ init_hw_perf_events(void) perf_max_events = armv7pmu.num_events; break; case 0xC090: /* Cortex-A9 */ - armv7pmu.name = ARMV7_PMU_CORTEX_A9_NAME; + armv7pmu.id = ARM_PERF_PMU_ID_CA9; memcpy(armpmu_perf_cache_map, armv7_a9_perf_cache_map, sizeof(armv7_a9_perf_cache_map)); armv7pmu.event_map = armv7_a9_pmu_event_map; @@ -2135,7 +2154,7 @@ init_hw_perf_events(void) if (armpmu) pr_info("enabled with %s PMU driver, %d counters available\n", - armpmu->name, armpmu->num_events); + arm_pmu_names[armpmu->id], armpmu->num_events); return 0; } -- 1.6.3.3 ^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 3/6] ARM: perf-events: add support for xscale PMUs 2010-02-25 18:56 ` [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU Will Deacon @ 2010-02-25 18:56 ` Will Deacon 2010-02-25 18:56 ` [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules Will Deacon 0 siblings, 1 reply; 19+ messages in thread From: Will Deacon @ 2010-02-25 18:56 UTC (permalink / raw) To: linux-arm-kernel The perf-events framework for ARM only supports v6 and v7 cores. This patch adds support for xscale v1 and v2 PMUs to perf, based on the OProfile drivers in arch/arm/oprofile/op_model_xscale.c Signed-off-by: Will Deacon <will.deacon@arm.com> --- arch/arm/kernel/perf_event.c | 825 +++++++++++++++++++++++++++++++++++++++++- 1 files changed, 819 insertions(+), 6 deletions(-) diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c index 51d9711..98c62ff 100644 --- a/arch/arm/kernel/perf_event.c +++ b/arch/arm/kernel/perf_event.c @@ -2097,6 +2097,801 @@ static u32 __init armv7_reset_read_pmnc(void) return nb_cnt + 1; } +/* + * ARMv5 [xscale] Performance counter handling code. + * + * Based on xscale OProfile code. + * + * There are two variants of the xscale PMU that we support: + * - xscale1pmu: 2 event counters and a cycle counter + * - xscale2pmu: 4 event counters and a cycle counter + * The two variants share event definitions, but have different + * PMU structures. + */ + +enum xscale_perf_types { + XSCALE_PERFCTR_ICACHE_MISS = 0x00, + XSCALE_PERFCTR_ICACHE_NO_DELIVER = 0x01, + XSCALE_PERFCTR_DATA_STALL = 0x02, + XSCALE_PERFCTR_ITLB_MISS = 0x03, + XSCALE_PERFCTR_DTLB_MISS = 0x04, + XSCALE_PERFCTR_BRANCH = 0x05, + XSCALE_PERFCTR_BRANCH_MISS = 0x06, + XSCALE_PERFCTR_INSTRUCTION = 0x07, + XSCALE_PERFCTR_DCACHE_FULL_STALL = 0x08, + XSCALE_PERFCTR_DCACHE_FULL_STALL_CONTIG = 0x09, + XSCALE_PERFCTR_DCACHE_ACCESS = 0x0A, + XSCALE_PERFCTR_DCACHE_MISS = 0x0B, + XSCALE_PERFCTR_DCACE_WRITE_BACK = 0x0C, + XSCALE_PERFCTR_PC_CHANGED = 0x0D, + XSCALE_PERFCTR_BCU_REQUEST = 0x10, + XSCALE_PERFCTR_BCU_FULL = 0x11, + XSCALE_PERFCTR_BCU_DRAIN = 0x12, + XSCALE_PERFCTR_BCU_ECC_NO_ELOG = 0x14, + XSCALE_PERFCTR_BCU_1_BIT_ERR = 0x15, + XSCALE_PERFCTR_RMW = 0x16, + /* XSCALE_PERFCTR_CCNT is not hardware defined */ + XSCALE_PERFCTR_CCNT = 0xFE, + XSCALE_PERFCTR_UNUSED = 0xFF, +}; + +enum xscale_counters { + XSCALE_CYCLE_COUNTER = 1, + XSCALE_COUNTER0, + XSCALE_COUNTER1, + XSCALE_COUNTER2, + XSCALE_COUNTER3, +}; + +static const unsigned xscale_perf_map[PERF_COUNT_HW_MAX] = { + [PERF_COUNT_HW_CPU_CYCLES] = XSCALE_PERFCTR_CCNT, + [PERF_COUNT_HW_INSTRUCTIONS] = XSCALE_PERFCTR_INSTRUCTION, + [PERF_COUNT_HW_CACHE_REFERENCES] = HW_OP_UNSUPPORTED, + [PERF_COUNT_HW_CACHE_MISSES] = HW_OP_UNSUPPORTED, + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = XSCALE_PERFCTR_BRANCH, + [PERF_COUNT_HW_BRANCH_MISSES] = XSCALE_PERFCTR_BRANCH_MISS, + [PERF_COUNT_HW_BUS_CYCLES] = HW_OP_UNSUPPORTED, +}; + +static const unsigned xscale_perf_cache_map[PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX] = { + [C(L1D)] = { + [C(OP_READ)] = { + [C(RESULT_ACCESS)] = XSCALE_PERFCTR_DCACHE_ACCESS, + [C(RESULT_MISS)] = XSCALE_PERFCTR_DCACHE_MISS, + }, + [C(OP_WRITE)] = { + [C(RESULT_ACCESS)] = XSCALE_PERFCTR_DCACHE_ACCESS, + [C(RESULT_MISS)] = XSCALE_PERFCTR_DCACHE_MISS, + }, + [C(OP_PREFETCH)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + }, + [C(L1I)] = { + [C(OP_READ)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = XSCALE_PERFCTR_ICACHE_MISS, + }, + [C(OP_WRITE)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = XSCALE_PERFCTR_ICACHE_MISS, + }, + [C(OP_PREFETCH)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + }, + [C(LL)] = { + [C(OP_READ)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + [C(OP_WRITE)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + [C(OP_PREFETCH)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + }, + [C(DTLB)] = { + [C(OP_READ)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = XSCALE_PERFCTR_DTLB_MISS, + }, + [C(OP_WRITE)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = XSCALE_PERFCTR_DTLB_MISS, + }, + [C(OP_PREFETCH)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + }, + [C(ITLB)] = { + [C(OP_READ)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = XSCALE_PERFCTR_ITLB_MISS, + }, + [C(OP_WRITE)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = XSCALE_PERFCTR_ITLB_MISS, + }, + [C(OP_PREFETCH)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + }, + [C(BPU)] = { + [C(OP_READ)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + [C(OP_WRITE)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + [C(OP_PREFETCH)] = { + [C(RESULT_ACCESS)] = CACHE_OP_UNSUPPORTED, + [C(RESULT_MISS)] = CACHE_OP_UNSUPPORTED, + }, + }, +}; + +#define XSCALE_PMU_ENABLE 0x001 +#define XSCALE_PMN_RESET 0x002 +#define XSCALE_CCNT_RESET 0x004 +#define XSCALE_PMU_RESET (CCNT_RESET | PMN_RESET) +#define XSCALE_PMU_CNT64 0x008 + +static inline int +xscalepmu_event_map(int config) +{ + int mapping = xscale_perf_map[config]; + if (HW_OP_UNSUPPORTED == mapping) + mapping = -EOPNOTSUPP; + return mapping; +} + +static u64 +xscalepmu_raw_event(u64 config) +{ + return config & 0xff; +} + +#define XSCALE1_OVERFLOWED_MASK 0x700 +#define XSCALE1_CCOUNT_OVERFLOW 0x400 +#define XSCALE1_COUNT0_OVERFLOW 0x100 +#define XSCALE1_COUNT1_OVERFLOW 0x200 +#define XSCALE1_CCOUNT_INT_EN 0x040 +#define XSCALE1_COUNT0_INT_EN 0x010 +#define XSCALE1_COUNT1_INT_EN 0x020 +#define XSCALE1_COUNT0_EVT_SHFT 12 +#define XSCALE1_COUNT0_EVT_MASK (0xff << XSCALE1_COUNT0_EVT_SHFT) +#define XSCALE1_COUNT1_EVT_SHFT 20 +#define XSCALE1_COUNT1_EVT_MASK (0xff << XSCALE1_COUNT1_EVT_SHFT) + +static inline u32 +xscale1pmu_read_pmnc(void) +{ + u32 val; + asm volatile("mrc p14, 0, %0, c0, c0, 0" : "=r" (val)); + return val; +} + +static inline void +xscale1pmu_write_pmnc(u32 val) +{ + /* upper 4bits and 7, 11 are write-as-0 */ + val &= 0xffff77f; + asm volatile("mcr p14, 0, %0, c0, c0, 0" : : "r" (val)); +} + +static inline int +xscale1_pmnc_counter_has_overflowed(unsigned long pmnc, + enum xscale_counters counter) +{ + int ret = 0; + + switch (counter) { + case XSCALE_CYCLE_COUNTER: + ret = pmnc & XSCALE1_CCOUNT_OVERFLOW; + break; + case XSCALE_COUNTER0: + ret = pmnc & XSCALE1_COUNT0_OVERFLOW; + break; + case XSCALE_COUNTER1: + ret = pmnc & XSCALE1_COUNT1_OVERFLOW; + break; + default: + WARN_ONCE(1, "invalid counter number (%d)\n", counter); + } + + return ret; +} + +static irqreturn_t +xscale1pmu_handle_irq(int irq_num, void *dev) +{ + unsigned long pmnc; + struct perf_sample_data data; + struct cpu_hw_events *cpuc; + struct pt_regs *regs; + int idx; + + /* + * NOTE: there's an A stepping erratum that states if an overflow + * bit already exists and another occurs, the previous + * Overflow bit gets cleared. There's no workaround. + * Fixed in B stepping or later. + */ + pmnc = xscale1pmu_read_pmnc(); + + /* + * Write the value back to clear the overflow flags. Overflow + * flags remain in pmnc for use below. We also disable the PMU + * while we process the interrupt. + */ + xscale1pmu_write_pmnc(pmnc & ~XSCALE_PMU_ENABLE); + + if (!(pmnc & XSCALE1_OVERFLOWED_MASK)) + return IRQ_NONE; + + regs = get_irq_regs(); + data.addr = 0; + + cpuc = &__get_cpu_var(cpu_hw_events); + for (idx = 0; idx <= armpmu->num_events; ++idx) { + struct perf_event *event = cpuc->events[idx]; + struct hw_perf_event *hwc; + + if (!test_bit(idx, cpuc->active_mask)) + continue; + + if (!xscale1_pmnc_counter_has_overflowed(pmnc, idx)) + continue; + + hwc = &event->hw; + armpmu_event_update(event, hwc, idx); + data.period = event->hw.last_period; + if (!armpmu_event_set_period(event, hwc, idx)) + continue; + + if (perf_event_overflow(event, 0, &data, regs)) + armpmu->disable(hwc, idx); + } + + perf_event_do_pending(); + + /* + * Re-enable the PMU. + */ + pmnc = xscale1pmu_read_pmnc() | XSCALE_PMU_ENABLE; + xscale1pmu_write_pmnc(pmnc); + + return IRQ_HANDLED; +} + +static void +xscale1pmu_enable_event(struct hw_perf_event *hwc, int idx) +{ + unsigned long val, mask, evt, flags; + + switch (idx) { + case XSCALE_CYCLE_COUNTER: + mask = 0; + evt = XSCALE1_CCOUNT_INT_EN; + break; + case XSCALE_COUNTER0: + mask = XSCALE1_COUNT0_EVT_MASK; + evt = (hwc->config_base << XSCALE1_COUNT0_EVT_SHFT) | + XSCALE1_COUNT0_INT_EN; + break; + case XSCALE_COUNTER1: + mask = XSCALE1_COUNT1_EVT_MASK; + evt = (hwc->config_base << XSCALE1_COUNT1_EVT_SHFT) | + XSCALE1_COUNT1_INT_EN; + break; + default: + WARN_ONCE(1, "invalid counter number (%d)\n", idx); + return; + } + + spin_lock_irqsave(&pmu_lock, flags); + val = xscale1pmu_read_pmnc(); + val &= ~mask; + val |= evt; + xscale1pmu_write_pmnc(val); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static void +xscale1pmu_disable_event(struct hw_perf_event *hwc, int idx) +{ + unsigned long val, mask, evt, flags; + + switch (idx) { + case XSCALE_CYCLE_COUNTER: + mask = XSCALE1_CCOUNT_INT_EN; + evt = 0; + break; + case XSCALE_COUNTER0: + mask = XSCALE1_COUNT0_INT_EN | XSCALE1_COUNT0_EVT_MASK; + evt = XSCALE_PERFCTR_UNUSED << XSCALE1_COUNT0_EVT_SHFT; + break; + case XSCALE_COUNTER1: + mask = XSCALE1_COUNT1_INT_EN | XSCALE1_COUNT1_EVT_MASK; + evt = XSCALE_PERFCTR_UNUSED << XSCALE1_COUNT1_EVT_SHFT; + break; + default: + WARN_ONCE(1, "invalid counter number (%d)\n", idx); + return; + } + + spin_lock_irqsave(&pmu_lock, flags); + val = xscale1pmu_read_pmnc(); + val &= ~mask; + val |= evt; + xscale1pmu_write_pmnc(val); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static int +xscale1pmu_get_event_idx(struct cpu_hw_events *cpuc, + struct hw_perf_event *event) +{ + if (XSCALE_PERFCTR_CCNT == event->config_base) { + if (test_and_set_bit(XSCALE_CYCLE_COUNTER, cpuc->used_mask)) + return -EAGAIN; + + return XSCALE_CYCLE_COUNTER; + } else { + if (!test_and_set_bit(XSCALE_COUNTER1, cpuc->used_mask)) { + return XSCALE_COUNTER1; + } + + if (!test_and_set_bit(XSCALE_COUNTER0, cpuc->used_mask)) { + return XSCALE_COUNTER0; + } + + return -EAGAIN; + } +} + +static void +xscale1pmu_start(void) +{ + unsigned long flags, val; + + spin_lock_irqsave(&pmu_lock, flags); + val = xscale1pmu_read_pmnc(); + val |= XSCALE_PMU_ENABLE; + xscale1pmu_write_pmnc(val); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static void +xscale1pmu_stop(void) +{ + unsigned long flags, val; + + spin_lock_irqsave(&pmu_lock, flags); + val = xscale1pmu_read_pmnc(); + val &= ~XSCALE_PMU_ENABLE; + xscale1pmu_write_pmnc(val); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static inline u32 +xscale1pmu_read_counter(int counter) +{ + u32 val = 0; + + switch (counter) { + case XSCALE_CYCLE_COUNTER: + asm volatile("mrc p14, 0, %0, c1, c0, 0" : "=r" (val)); + break; + case XSCALE_COUNTER0: + asm volatile("mrc p14, 0, %0, c2, c0, 0" : "=r" (val)); + break; + case XSCALE_COUNTER1: + asm volatile("mrc p14, 0, %0, c3, c0, 0" : "=r" (val)); + break; + } + + return val; +} + +static inline void +xscale1pmu_write_counter(int counter, u32 val) +{ + switch (counter) { + case XSCALE_CYCLE_COUNTER: + asm volatile("mcr p14, 0, %0, c1, c0, 0" : : "r" (val)); + break; + case XSCALE_COUNTER0: + asm volatile("mcr p14, 0, %0, c2, c0, 0" : : "r" (val)); + break; + case XSCALE_COUNTER1: + asm volatile("mcr p14, 0, %0, c3, c0, 0" : : "r" (val)); + break; + } +} + +static const struct arm_pmu xscale1pmu = { + .id = ARM_PERF_PMU_ID_XSCALE1, + .handle_irq = xscale1pmu_handle_irq, + .enable = xscale1pmu_enable_event, + .disable = xscale1pmu_disable_event, + .event_map = xscalepmu_event_map, + .raw_event = xscalepmu_raw_event, + .read_counter = xscale1pmu_read_counter, + .write_counter = xscale1pmu_write_counter, + .get_event_idx = xscale1pmu_get_event_idx, + .start = xscale1pmu_start, + .stop = xscale1pmu_stop, + .num_events = 3, + .max_period = (1LLU << 32) - 1, +}; + +#define XSCALE2_OVERFLOWED_MASK 0x01f +#define XSCALE2_CCOUNT_OVERFLOW 0x001 +#define XSCALE2_COUNT0_OVERFLOW 0x002 +#define XSCALE2_COUNT1_OVERFLOW 0x004 +#define XSCALE2_COUNT2_OVERFLOW 0x008 +#define XSCALE2_COUNT3_OVERFLOW 0x010 +#define XSCALE2_CCOUNT_INT_EN 0x001 +#define XSCALE2_COUNT0_INT_EN 0x002 +#define XSCALE2_COUNT1_INT_EN 0x004 +#define XSCALE2_COUNT2_INT_EN 0x008 +#define XSCALE2_COUNT3_INT_EN 0x010 +#define XSCALE2_COUNT0_EVT_SHFT 0 +#define XSCALE2_COUNT0_EVT_MASK (0xff << XSCALE2_COUNT0_EVT_SHFT) +#define XSCALE2_COUNT1_EVT_SHFT 8 +#define XSCALE2_COUNT1_EVT_MASK (0xff << XSCALE2_COUNT1_EVT_SHFT) +#define XSCALE2_COUNT2_EVT_SHFT 16 +#define XSCALE2_COUNT2_EVT_MASK (0xff << XSCALE2_COUNT2_EVT_SHFT) +#define XSCALE2_COUNT3_EVT_SHFT 24 +#define XSCALE2_COUNT3_EVT_MASK (0xff << XSCALE2_COUNT3_EVT_SHFT) + +static inline u32 +xscale2pmu_read_pmnc(void) +{ + u32 val; + asm volatile("mrc p14, 0, %0, c0, c1, 0" : "=r" (val)); + /* bits 1-2 and 4-23 are read-unpredictable */ + return val & 0xff000009; +} + +static inline void +xscale2pmu_write_pmnc(u32 val) +{ + /* bits 4-23 are write-as-0, 24-31 are write ignored */ + val &= 0xf; + asm volatile("mcr p14, 0, %0, c0, c1, 0" : : "r" (val)); +} + +static inline u32 +xscale2pmu_read_overflow_flags(void) +{ + u32 val; + asm volatile("mrc p14, 0, %0, c5, c1, 0" : "=r" (val)); + return val; +} + +static inline void +xscale2pmu_write_overflow_flags(u32 val) +{ + asm volatile("mcr p14, 0, %0, c5, c1, 0" : : "r" (val)); +} + +static inline u32 +xscale2pmu_read_event_select(void) +{ + u32 val; + asm volatile("mrc p14, 0, %0, c8, c1, 0" : "=r" (val)); + return val; +} + +static inline void +xscale2pmu_write_event_select(u32 val) +{ + asm volatile("mcr p14, 0, %0, c8, c1, 0" : : "r"(val)); +} + +static inline u32 +xscale2pmu_read_int_enable(void) +{ + u32 val; + asm volatile("mrc p14, 0, %0, c4, c1, 0" : "=r" (val)); + return val; +} + +static void +xscale2pmu_write_int_enable(u32 val) +{ + asm volatile("mcr p14, 0, %0, c4, c1, 0" : : "r" (val)); +} + +static inline int +xscale2_pmnc_counter_has_overflowed(unsigned long of_flags, + enum xscale_counters counter) +{ + int ret = 0; + + switch (counter) { + case XSCALE_CYCLE_COUNTER: + ret = of_flags & XSCALE2_CCOUNT_OVERFLOW; + break; + case XSCALE_COUNTER0: + ret = of_flags & XSCALE2_COUNT0_OVERFLOW; + break; + case XSCALE_COUNTER1: + ret = of_flags & XSCALE2_COUNT1_OVERFLOW; + break; + case XSCALE_COUNTER2: + ret = of_flags & XSCALE2_COUNT2_OVERFLOW; + break; + case XSCALE_COUNTER3: + ret = of_flags & XSCALE2_COUNT3_OVERFLOW; + break; + default: + WARN_ONCE(1, "invalid counter number (%d)\n", counter); + } + + return ret; +} + +static irqreturn_t +xscale2pmu_handle_irq(int irq_num, void *dev) +{ + unsigned long pmnc, of_flags; + struct perf_sample_data data; + struct cpu_hw_events *cpuc; + struct pt_regs *regs; + int idx; + + /* Disable the PMU. */ + pmnc = xscale2pmu_read_pmnc(); + xscale2pmu_write_pmnc(pmnc & ~XSCALE_PMU_ENABLE); + + /* Check the overflow flag register. */ + of_flags = xscale2pmu_read_overflow_flags(); + if (!(of_flags & XSCALE2_OVERFLOWED_MASK)) + return IRQ_NONE; + + /* Clear the overflow bits. */ + xscale2pmu_write_overflow_flags(of_flags); + + regs = get_irq_regs(); + data.addr = 0; + + cpuc = &__get_cpu_var(cpu_hw_events); + for (idx = 0; idx <= armpmu->num_events; ++idx) { + struct perf_event *event = cpuc->events[idx]; + struct hw_perf_event *hwc; + + if (!test_bit(idx, cpuc->active_mask)) + continue; + + if (!xscale2_pmnc_counter_has_overflowed(pmnc, idx)) + continue; + + hwc = &event->hw; + armpmu_event_update(event, hwc, idx); + data.period = event->hw.last_period; + if (!armpmu_event_set_period(event, hwc, idx)) + continue; + + if (perf_event_overflow(event, 0, &data, regs)) + armpmu->disable(hwc, idx); + } + + perf_event_do_pending(); + + /* + * Re-enable the PMU. + */ + pmnc = xscale2pmu_read_pmnc() | XSCALE_PMU_ENABLE; + xscale2pmu_write_pmnc(pmnc); + + return IRQ_HANDLED; +} + +static void +xscale2pmu_enable_event(struct hw_perf_event *hwc, int idx) +{ + unsigned long flags, ien, evtsel; + + ien = xscale2pmu_read_int_enable(); + evtsel = xscale2pmu_read_event_select(); + + switch (idx) { + case XSCALE_CYCLE_COUNTER: + ien |= XSCALE2_CCOUNT_INT_EN; + break; + case XSCALE_COUNTER0: + ien |= XSCALE2_COUNT0_INT_EN; + evtsel &= ~XSCALE2_COUNT0_EVT_MASK; + evtsel |= hwc->config_base << XSCALE2_COUNT0_EVT_SHFT; + break; + case XSCALE_COUNTER1: + ien |= XSCALE2_COUNT1_INT_EN; + evtsel &= ~XSCALE2_COUNT1_EVT_MASK; + evtsel |= hwc->config_base << XSCALE2_COUNT1_EVT_SHFT; + break; + case XSCALE_COUNTER2: + ien |= XSCALE2_COUNT2_INT_EN; + evtsel &= ~XSCALE2_COUNT2_EVT_MASK; + evtsel |= hwc->config_base << XSCALE2_COUNT2_EVT_SHFT; + break; + case XSCALE_COUNTER3: + ien |= XSCALE2_COUNT3_INT_EN; + evtsel &= ~XSCALE2_COUNT3_EVT_MASK; + evtsel |= hwc->config_base << XSCALE2_COUNT3_EVT_SHFT; + break; + default: + WARN_ONCE(1, "invalid counter number (%d)\n", idx); + return; + } + + spin_lock_irqsave(&pmu_lock, flags); + xscale2pmu_write_event_select(evtsel); + xscale2pmu_write_int_enable(ien); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static void +xscale2pmu_disable_event(struct hw_perf_event *hwc, int idx) +{ + unsigned long flags, ien, evtsel; + + ien = xscale2pmu_read_int_enable(); + evtsel = xscale2pmu_read_event_select(); + + switch (idx) { + case XSCALE_CYCLE_COUNTER: + ien &= ~XSCALE2_CCOUNT_INT_EN; + break; + case XSCALE_COUNTER0: + ien &= ~XSCALE2_COUNT0_INT_EN; + evtsel &= ~XSCALE2_COUNT0_EVT_MASK; + evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT0_EVT_SHFT; + break; + case XSCALE_COUNTER1: + ien &= ~XSCALE2_COUNT1_INT_EN; + evtsel &= ~XSCALE2_COUNT1_EVT_MASK; + evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT1_EVT_SHFT; + break; + case XSCALE_COUNTER2: + ien &= ~XSCALE2_COUNT2_INT_EN; + evtsel &= ~XSCALE2_COUNT2_EVT_MASK; + evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT2_EVT_SHFT; + break; + case XSCALE_COUNTER3: + ien &= ~XSCALE2_COUNT3_INT_EN; + evtsel &= ~XSCALE2_COUNT3_EVT_MASK; + evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT3_EVT_SHFT; + break; + default: + WARN_ONCE(1, "invalid counter number (%d)\n", idx); + return; + } + + spin_lock_irqsave(&pmu_lock, flags); + xscale2pmu_write_event_select(evtsel); + xscale2pmu_write_int_enable(ien); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static int +xscale2pmu_get_event_idx(struct cpu_hw_events *cpuc, + struct hw_perf_event *event) +{ + int idx = xscale1pmu_get_event_idx(cpuc, event); + if (idx >= 0) + goto out; + + if (!test_and_set_bit(XSCALE_COUNTER3, cpuc->used_mask)) + idx = XSCALE_COUNTER3; + else if (!test_and_set_bit(XSCALE_COUNTER2, cpuc->used_mask)) + idx = XSCALE_COUNTER2; +out: + return idx; +} + +static void +xscale2pmu_start(void) +{ + unsigned long flags, val; + + spin_lock_irqsave(&pmu_lock, flags); + val = xscale2pmu_read_pmnc() & ~XSCALE_PMU_CNT64; + val |= XSCALE_PMU_ENABLE; + xscale2pmu_write_pmnc(val); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static void +xscale2pmu_stop(void) +{ + unsigned long flags, val; + + spin_lock_irqsave(&pmu_lock, flags); + val = xscale2pmu_read_pmnc(); + val &= ~XSCALE_PMU_ENABLE; + xscale2pmu_write_pmnc(val); + spin_unlock_irqrestore(&pmu_lock, flags); +} + +static inline u32 +xscale2pmu_read_counter(int counter) +{ + u32 val = 0; + + switch (counter) { + case XSCALE_CYCLE_COUNTER: + asm volatile("mrc p14, 0, %0, c1, c1, 0" : "=r" (val)); + break; + case XSCALE_COUNTER0: + asm volatile("mrc p14, 0, %0, c0, c2, 0" : "=r" (val)); + break; + case XSCALE_COUNTER1: + asm volatile("mrc p14, 0, %0, c1, c2, 0" : "=r" (val)); + break; + case XSCALE_COUNTER2: + asm volatile("mrc p14, 0, %0, c2, c2, 0" : "=r" (val)); + break; + case XSCALE_COUNTER3: + asm volatile("mrc p14, 0, %0, c3, c2, 0" : "=r" (val)); + break; + } + + return val; +} + +static inline void +xscale2pmu_write_counter(int counter, u32 val) +{ + switch (counter) { + case XSCALE_CYCLE_COUNTER: + asm volatile("mcr p14, 0, %0, c1, c1, 0" : : "r" (val)); + break; + case XSCALE_COUNTER0: + asm volatile("mcr p14, 0, %0, c0, c2, 0" : : "r" (val)); + break; + case XSCALE_COUNTER1: + asm volatile("mcr p14, 0, %0, c1, c2, 0" : : "r" (val)); + break; + case XSCALE_COUNTER2: + asm volatile("mcr p14, 0, %0, c2, c2, 0" : : "r" (val)); + break; + case XSCALE_COUNTER3: + asm volatile("mcr p14, 0, %0, c3, c2, 0" : : "r" (val)); + break; + } +} + +static const struct arm_pmu xscale2pmu = { + .id = ARM_PERF_PMU_ID_XSCALE2, + .handle_irq = xscale2pmu_handle_irq, + .enable = xscale2pmu_enable_event, + .disable = xscale2pmu_disable_event, + .event_map = xscalepmu_event_map, + .raw_event = xscalepmu_raw_event, + .read_counter = xscale2pmu_read_counter, + .write_counter = xscale2pmu_write_counter, + .get_event_idx = xscale2pmu_get_event_idx, + .start = xscale2pmu_start, + .stop = xscale2pmu_stop, + .num_events = 5, + .max_period = (1LLU << 32) - 1, +}; + static int __init init_hw_perf_events(void) { @@ -2104,7 +2899,7 @@ init_hw_perf_events(void) unsigned long implementor = (cpuid & 0xFF000000) >> 24; unsigned long part_number = (cpuid & 0xFFF0); - /* We only support ARM CPUs implemented by ARM at the moment. */ + /* ARM Ltd CPUs. */ if (0x41 == implementor) { switch (part_number) { case 0xB360: /* ARM1136 */ @@ -2146,15 +2941,33 @@ init_hw_perf_events(void) armv7pmu.num_events = armv7_reset_read_pmnc(); perf_max_events = armv7pmu.num_events; break; - default: - pr_info("no hardware support available\n"); - perf_max_events = -1; + } + /* Intel CPUs [xscale]. */ + } else if (0x69 == implementor) { + part_number = (cpuid >> 13) & 0x7; + switch (part_number) { + case 1: + armpmu = &xscale1pmu; + memcpy(armpmu_perf_cache_map, xscale_perf_cache_map, + sizeof(xscale_perf_cache_map)); + perf_max_events = xscale1pmu.num_events; + break; + case 2: + armpmu = &xscale2pmu; + memcpy(armpmu_perf_cache_map, xscale_perf_cache_map, + sizeof(xscale_perf_cache_map)); + perf_max_events = xscale2pmu.num_events; + break; } } - if (armpmu) + if (armpmu) { pr_info("enabled with %s PMU driver, %d counters available\n", - arm_pmu_names[armpmu->id], armpmu->num_events); + arm_pmu_names[armpmu->id], armpmu->num_events); + } else { + pr_info("no hardware support available\n"); + perf_max_events = -1; + } return 0; } -- 1.6.3.3 ^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules 2010-02-25 18:56 ` [PATCH 3/6] ARM: perf-events: add support for xscale PMUs Will Deacon @ 2010-02-25 18:56 ` Will Deacon 2010-02-26 8:23 ` Ingo Molnar 0 siblings, 1 reply; 19+ messages in thread From: Will Deacon @ 2010-02-25 18:56 UTC (permalink / raw) To: linux-arm-kernel The perf_event_enable and perf_event_disable functions are used to control the activation of perf-events for a given context or CPU. This patch exports these symbols so that they can be used by kernel modules. Without these symbols, an event must be destroyed and recreated to disable or enable it respectively. The maximum number of perf-events is also made available to modules via the perf_get_max_events function. Cc: Ingo Molnar <mingo@elte.hu> Signed-off-by: Will Deacon <will.deacon@arm.com> --- include/linux/perf_event.h | 1 + kernel/perf_event.c | 8 ++++++++ 2 files changed, 9 insertions(+), 0 deletions(-) diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index a177698..c28a0ee 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -741,6 +741,7 @@ struct perf_output_handle { * Set by architecture code: */ extern int perf_max_events; +extern int perf_get_max_events(void); extern const struct pmu *hw_perf_event_init(struct perf_event *event); diff --git a/kernel/perf_event.c b/kernel/perf_event.c index 2ae7409..0157033 100644 --- a/kernel/perf_event.c +++ b/kernel/perf_event.c @@ -39,6 +39,12 @@ static DEFINE_PER_CPU(struct perf_cpu_context, perf_cpu_context); int perf_max_events __read_mostly = 1; +int perf_get_max_events(void) +{ + return perf_max_events; +} +EXPORT_SYMBOL_GPL(perf_get_max_events); + static int perf_reserved_percpu __read_mostly; static int perf_overcommit __read_mostly = 1; @@ -604,6 +610,7 @@ void perf_event_disable(struct perf_event *event) raw_spin_unlock_irq(&ctx->lock); } +EXPORT_SYMBOL_GPL(perf_event_disable); static int event_sched_in(struct perf_event *event, @@ -1028,6 +1035,7 @@ void perf_event_enable(struct perf_event *event) out: raw_spin_unlock_irq(&ctx->lock); } +EXPORT_SYMBOL_GPL(perf_event_enable); static int perf_event_refresh(struct perf_event *event, int refresh) { -- 1.6.3.3 ^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules 2010-02-25 18:56 ` [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules Will Deacon @ 2010-02-26 8:23 ` Ingo Molnar 2010-02-26 9:53 ` Will Deacon 0 siblings, 1 reply; 19+ messages in thread From: Ingo Molnar @ 2010-02-26 8:23 UTC (permalink / raw) To: linux-arm-kernel * Will Deacon <will.deacon@arm.com> wrote: > The perf_event_enable and perf_event_disable functions are used to > control the activation of perf-events for a given context or CPU. > > This patch exports these symbols so that they can be used by kernel > modules. Without these symbols, an event must be destroyed and recreated > to disable or enable it respectively. The maximum number of perf-events > is also made available to modules via the perf_get_max_events function. > > Cc: Ingo Molnar <mingo@elte.hu> > Signed-off-by: Will Deacon <will.deacon@arm.com> > --- > include/linux/perf_event.h | 1 + > kernel/perf_event.c | 8 ++++++++ > 2 files changed, 9 insertions(+), 0 deletions(-) Hm, what modules would like to use them, and in what fashion? Thanks, Ingo ^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules 2010-02-26 8:23 ` Ingo Molnar @ 2010-02-26 9:53 ` Will Deacon 0 siblings, 0 replies; 19+ messages in thread From: Will Deacon @ 2010-02-26 9:53 UTC (permalink / raw) To: linux-arm-kernel Hi Ingo, Thanks for the quick response. > * Will Deacon <will.deacon@arm.com> wrote: > > > The perf_event_enable and perf_event_disable functions are used to > > control the activation of perf-events for a given context or CPU. > > > > This patch exports these symbols so that they can be used by kernel > > modules. Without these symbols, an event must be destroyed and recreated > > to disable or enable it respectively. The maximum number of perf-events > > is also made available to modules via the perf_get_max_events function. > > > > Hm, what modules would like to use them, and in what fashion? > The module I had in mind was OProfile [currently just the ARM port, but perhaps other architectures may decide to do this too]. The patch I posted here: http://lists.infradead.org/pipermail/linux-arm-kernel/2010-February/010479.html removes the PMU control code from OProfile and replaces it with calls to perf. This means that when OProfile starts or stops a profiling session, the corresponding events are enabled or disabled respectively. If you want to compile OProfile as a module, then the enable/disable functions will need to be exported. I suppose the other solution is to prevent OProfile from being compiled as a module. Thanks, Will ^ permalink raw reply [flat|nested] 19+ messages in thread
end of thread, other threads:[~2010-03-10 22:34 UTC | newest] Thread overview: 19+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2010-03-10 10:41 [PATCH 0/6] ARM: oprofile: use perf-events framework as backend [v2] Will Deacon 2010-03-10 10:41 ` [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c Will Deacon 2010-03-10 10:41 ` [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU Will Deacon 2010-03-10 10:41 ` [PATCH 3/6] ARM: perf-events: add support for xscale PMUs Will Deacon 2010-03-10 10:41 ` [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules Will Deacon 2010-03-10 10:41 ` [PATCH 5/6] ARM: oprofile: use perf-events framework as backend Will Deacon 2010-03-10 10:41 ` [PATCH 6/6] ARM: oprofile: remove old files and update KConfig Will Deacon 2010-03-10 10:56 ` [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules Peter Zijlstra 2010-03-10 11:47 ` Will Deacon 2010-03-10 14:30 ` Peter Zijlstra 2010-03-10 14:41 ` Will Deacon 2010-03-10 15:01 ` Robert Richter 2010-03-10 11:03 ` [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU Jamie Iles 2010-03-10 21:59 ` [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c Tony Lindgren 2010-03-10 22:27 ` Albin Tonnerre 2010-03-10 22:34 ` Tony Lindgren -- strict thread matches above, loose matches on Subject: below -- 2010-02-25 18:56 [PATCH 0/6] ARM: oprofile: use perf-events framework as backend Will Deacon 2010-02-25 18:56 ` [PATCH 1/6] ARM: perf-events: add Realview PMU IRQs to pmu.c Will Deacon 2010-02-25 18:56 ` [PATCH 2/6] ARM: perf-events: use numeric ID to identify PMU Will Deacon 2010-02-25 18:56 ` [PATCH 3/6] ARM: perf-events: add support for xscale PMUs Will Deacon 2010-02-25 18:56 ` [PATCH 4/6] perf-events: export enable/disable event symbols to kernel modules Will Deacon 2010-02-26 8:23 ` Ingo Molnar 2010-02-26 9:53 ` Will Deacon
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).